How does a machine know?

I'm curious, maybe the board experts can advise...

How does my machine know the difference between types of events? All these events are concerned with a stop of breathing in one way or another: central vs. obstructive vs. Hypopnea. How does it tell the difference between a sinus event to a throat event? And while we're on the subject, how does it measure snoring? Are there sensors in the machine that I'm not aware of?

Are you interested in a list of articles and reports that explain all these? But first please reply how could I tell the answer to your question about those sensors that you don't know about?

Overall, yes, there are sensors that measure most of the following parameters, but notice that while in Sleep Labs all of these are usually measured and recorded only those marked with a star are measured in home- CPAPs and an Oximeter:

1) Airflow* and Pressures*
2) Snore*
3) Plethysmography (some time)
4) Heart Rate*
5) Chest Effort
6) Abdominal Effort
7) Body Position
Mask Pressure* (in some)
9) Mask Leak*
10) SpO2*
11) Limb Movements
12) EKG
13) EEG
14) EMG of chin and tibials

As to a comparison between measures and calculations between in-home-CPAPS and those at Sleep Study centers:
(as of June 2010)

Studies that compared CPAP-derived AHI to in-lab PSG, AHI showed marked differences between the two concluding that CPAP’s have insufficient signal quality and could potentially generate” misleading results” especially for mild to moderate OSA patients. These studies demonstrate that while superior to oximetry alone and therefore a potentially better in-home “screening” tool, CPAP’s without supplemental information can overestimate apneas and underestimate hypopneas. The studies also showed susceptibility to mouth leaks and mask motion artifacts.

The short answer is that each type of event has a different profile.

For example, a snore exhibits a "vibration" of the flow resistance during inhale. Other types of events have signatures that involve different phases of the event. But you should know that your machine doesn't always get it right.

LinkC said: your machine doesn't always get it right

Do you have an opinion as to how much better the equipment is that is used in a Sleep Study?

LinkC wrote: your machine doesn't always get it right

I'd rather that it not always get it right but at least be there to help me through the episode than be 100% all of the time and be hit-or-miss as far as reliability/efficacy are concerned... an episode is an episode is an episode and as long as I don't go back to how I was on the night of December 12, 2010 (the night before the start of therapy), then a miss diagnosis of what type of event I have is really nothing to get bent out of shape about.

All right, thank you for the responses. I understand how sleep centers could get that information because there are sensors all over the place. I was curious about how our home machines might derive that information. For example, I'm using an S9 Auto and notice on my reports that some events are classified differently than others. So was curious as to how the S9 determines which is which...

You might want to take a look at the clinician's manual for the S9 (google should find it). It explains how the S9 identifies each event as well as the S9's response to each type of event.

Good tip, thank you!

You can get the Clinian Manual for the ResMed S9 AutoSet and all of the others in the S9 Series by following the directions below. All of the models in the S9 Series are in the same Clinian Manual.

We have a few CPAP Clinician/Setup Manuals (that are too large to post on this web site, but are available through email.

Please follow these email instructions exactly:

If you would like one of the manuals below, please send an email to: apneaboard@gmail.com and put "Manuals" (without the quotes) in the subject line. Then, let us know in the body of your email which of the following manuals you would like us to email to you.

LIMIT: Two (2) Clinician Manuals per individual.
(If you request more than two manuals, your request will be discarded)

I tried to PM the Clinician Manual to you and even tried to attach it in this reply - but apparently the file was too large.

Resmed's website has some information on their central apnea detection that you might find interesting:
http://www.resmed.com/us/products/s9_se ... nc=dealers

"It uses the forced oscillation technique (FOT) to determine the state of the airway during an apnea. When an apnea is detected, small oscillations are added to the pressure to measure airway patency. The CSA algorithm uses the resulting flow and pressure to measure airway patency and differentiate central and obstructive events."

Hephaestus wrote:All right, thank you for the responses. I understand how sleep centers could get that information because there are sensors all over the place. I was curious about how our home machines might derive that information. For example, I'm using an S9 Auto and notice on my reports that some events are classified differently than others. So was curious as to how the S9 determines which is which...

The S9 has computer chips inside similar to a laptop. It also has lots of tiny sensors to measure the physical parameters.

I believe xPAPs generally have just a single type of sensor which measures only airflow*. As some prior posters mention, by measuring the "baseline" flow and the resulting flow when certain variations are purposely introduced by the machine, the machine is able to classify events.

* Maybe some also measure temp and humidity, but these I think are mainly for comfort, not for diagnosis.

avi123 wrote:
As to a comparison between measures and calculations between in-home-CPAPS and those at Sleep Study centers:
(as of June 2010)

Studies that compared CPAP-derived AHI to in-lab PSG, AHI showed marked differences between the two concluding that CPAP’s have insufficient signal quality and could potentially generate” misleading results” especially for mild to moderate OSA patients. These studies demonstrate that while superior to oximetry alone and therefore a potentially better in-home “screening” tool, CPAP’s without supplemental information can overestimate apneas and underestimate hypopneas. The studies also showed susceptibility to mouth leaks and mask motion artifacts.

avi123,

Could you provide a source for the quote you posted?

Thanks.

Midnightowl

jmelby wrote:Resmed's website has some information on their central apnea detection that you might find interesting:
http://www.resmed.com/us/products/s9_se ... nc=dealers

"It uses the forced oscillation technique (FOT) to determine the state of the airway during an apnea. When an apnea is detected, small oscillations are added to the pressure to measure airway patency. The CSA algorithm uses the resulting flow and pressure to measure airway patency and differentiate central and obstructive events."

Very interesting! Knowing this tells me that our modern machines are much more than just a simple air pump. I wouldn't have known that we could gather this type of information via air pressure/flow measurements. Certainly we couldn't do it without cpu's and software algorithm's to precisely control the air pressure and volume.

I'd always thought it would be a good move to get into the medical equipment field!