Distorted or misrepresented source data OSCAR vs. SleepHQ

[dataq1]

In another thread viewtopic/t185468/Pulse-Oximeters-in-CP ... 0#p1426647 (as a somewhat off-topic comment):

It should be quite clear that the Mask Pressure graph from SleepHQ is not displaying (all) the data that was written on the SD card correctly.

If SleepHQ is incorrectly displaying primary source (measured) data, that ought to be discussed.

[chunkyfrog]

Maybe they can fix that problem, too.
It must still be beta . . .

[Rubicon]

One likes to believe in the freedom of music; but glittering prizes and endless compromises shatter the illusion of integrity.

[lazarus]

"Mess and magic, triumphant and tragic
A mechanized world out of hand
Computerized clinic for superior cynics
Who dance to a synthetic band"

--Natural Science, lyrics by N. Peart.

[Wanderlustralia]

The browsers would crash in an instant if they were hit with super high res data with hundreds of thousands of data points. It would cause all sorts of performance issues. The data is the same, just lower res. It's the same thing youtube does when you're on a slow internet speed. Instead of showing you the 4k stuff and buffering etc, they show you the 720p. For 99.9% of viewers, no one cares! The same goes for SleepHQ. For 99.9% of the users. They're happy with the level of detail provided. SleepHQ is not OSCAR! It was made for the average Joe.

[Rubicon]

"Mess and magic, triumphant and tragic
A mechanized world out of hand
Computerized clinic for superior cynics
Who dance to a synthetic band"

--Natural Science, lyrics by N. Peart.

I wonder if Geddy uses a teleprompter...

[lazarus]

Not in the old days. But purdy much everyone in more modern times:

https://www.youtube.com/watch?v=VEF320HwGYs#t=376

[Rubicon]

I'd like to put the lyrics for "Positively 4th Street" on the teleprompter for when he does "Spirit" to see what happens.

Bet he wouldn't miss a beat.

Literally.

[lynninnj]

Not in the old days. But purdy much everyone in more modern times:

https://www.youtube.com/watch?v=VEF320HwGYs#t=376

great interview thx

RIP Neal Peart

[dataq1]

The browsers would crash in an instant if they were hit with super high res data with hundreds of thousands of data points. It would cause all sorts of performance issues. The data is the same, just lower res. It's the same thing youtube does when you're on a slow internet speed. Instead of showing you the 4k stuff and buffering etc, they show you the 720p. For 99.9% of viewers, no one cares! The same goes for SleepHQ. For 99.9% of the users. They're happy with the level of detail provided. SleepHQ is not OSCAR! It was made for the average Joe.

A completely reasonable explanation, however I can't help but wonder if you (Wanderlustralia) are the official voice of SleepHQ?
If so, then it is significant that the "official voice" of SleepHQ is telling the community that SleepHQ should not be directly compared with OSCAR, because (for completely practical purposes) SleepHQ is examining the raw (primary source) data at a reduced resolution.

[robysue1]

The browsers would crash in an instant if they were hit with super high res data with hundreds of thousands of data points. It would cause all sorts of performance issues. The data is the same, just lower res.

Then why don't the browsers crash in an instant when they are hit with the flow rate data? The flow rate data is also super high res data with hundreds of thousands of data points as the machine tracks the flow rate breath-by-breath, all night long.

The differences between the Oscar flow rate graph and the SleepHQ flow rate graph are minimal: The only one I can actually find is that you can't reset the y-axis scale in SleepHQ, but you can in Oscar.

A properly drawn mask pressure graph shows the pressure changes between IPAP and EPAP on each and every breath. The SleepHQ mask pressure graph does not show how Mask Pressure is directly related to the individual breaths in the Flow Rate graph, and this is particularly evident when you zoom in on just a hand full of breaths.

Here's a 30-second snippet of pretty decent sleep breathing from my Nov. 14 data as shown in Oscar. (The vertical scale on the Flow Rate graph has been chosen to correspond to the range picked by SleepHQ.)

You can see how the Mask Pressure indicates exactly when the changes between IPAP = 8 and EPAP = 4 are occurring. In other words, the Mask Pressure confirms the machine is indeed doing its job of switching the pressures in sync with my breathing.

Here's the same 30-second snippet of pretty decent sleep breathing from my Nov. 14 data as shown in SleepHQ:


You can see that the flow rate curves look the same. In other words, SleepHQ is using high res data to draw that curve correctly. But that mask pressure curve? You can't see any of the EasyBreathe algorithm's method of transitioning the EPAP pressure from 4cm to the IPAP pressure of 8cm. If you believe this graph, it implies my pressure is pretty constant at 5cm throughout this 30 second snippet.

Of course, it's possible to say that "lo res" data shouldn't be zoomed in to 30-second epochs. But even if we zoom all the way to look at the whole night's data, you can tell there's something wrong with the mask pressure graph. Here's the correctly drawn flow rate and mask pressure graph for the whole night in Oscar:


Here's the same data in SleepHQ with the same size graphs and the same scale on the flow rate graph:



It's clear that the Oscar graph is correctly picking up on the fact that my mask pressure is going from 4cm (on each exhalation) to 8cm (by the end of each inhalation) at that long stretch at the beginning of the night. SleepHQ's graph makes it look like my mask pressure is ranging from about 4.8cm to about 5.25cm during that same stretch. That's a pretty big "error" if you ask me. And it totally hides what this graph is supposed to show---which is how the PS = 4 is being implemented on a breath by breath basis.

It's the same thing youtube does when you're on a slow internet speed. Instead of showing you the 4k stuff and buffering etc, they show you the 720p. For 99.9% of viewers, no one cares! The same goes for SleepHQ. For 99.9% of the users. They're happy with the level of detail provided. SleepHQ is not OSCAR! It was made for the average Joe.

Are you saying that the designers of SleepHQ made a decision to intentionally present a low res presentation of data that is, by its nature, high res, on the mask pressure curve because they though the average user wouldn't notice or wouldn't care? But they also chose to keep the high res presentation for the flow rate graph because that was something even the most casual user would notice was wrong if it was presented in a low res format?

[lazarus]

My assumption (and it is nothing more than that, since I have no direct knowledge of the matter) is that OSCAR can legally get away with providing more direct representation of what the machine itself is doing because of the decidedly nonprofit nature of the project.

I would think that any project that plans to go for-profit would have to be much more careful with presenting proprietary information about the machine's actions than it would need to be in presenting patient actions and response, since it can be argued that the patient owns the data regarding his or her own breathing and responses.

A split of a hair? Maybe. But that's what many business lawyers get paid to do: quantify legal risks with product design and features.

Then again, I may be wrong. I often am.

And cutting data by halving the amount of high-resolution detail that is stored, overall, may be the economic consideration entirely.

[lazarus]

great interview thx

RIP Neal Peart

A master interviewer speaking to The Professor himself:

https://youtu.be/q_mKr28G7og

[dataq1]

robysue1:
In your first two sets of comparison graphs, the time is off by one minute ( OSCAR starts at 00:00:00 and SleepHQ starts at 12:01:00 ) and the Yaxis for the Mask Pressure is not the same between OSCAR and SleepHQ. Assuming that your breathing did not change over the course of 60 seconds, your point is still very well taken, that the SleepHQ mask graph is "flat" as compared with the OSCAR mask graph. If you changed the Yaxis the SleepHQ graph would be even flatter !

[Wanderlustralia]

The info i provided was accurate. I'm not here for a debate. Cheers