+- 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: Why don't these flow rate curves count as flow-limited? (/Thread-Why-don-t-these-flow-rate-curves-count-as-flow-limited)
RE: Why don't these flow rate curves count as flow-limited? - cathyf - 07-18-2021
(07-16-2021, 02:33 AM)multicast Wrote: Well, not "really": A real, physical flow limitation is something that happens with your body, it's not about graphs. A happening flow limitation is a non-linear response
between your breathing efforts and the amount of air you're actually breathe. That is: your belly is moving up and down but there's only few air reaching your lounges. It's not possible to detect a real FL with our machines for certain.
Let me restate that...
"As an aside, I think I now understand what detecting a flow limit is about. It's NOT about looking at the tippy-tops of those inspiratory flow rate waves like entrails of sacrificial goats. It's all about how the area is distributed under the curves."
RE: Why don't these flow rate curves count as flow-limited? - multicast - 07-18-2021
(07-18-2021, 02:30 AM)cathyf Wrote: "It's all about how the area is distributed under the curves."
In a way ... not. You've read that mention patent https://patents.google.com/patent/US20110203588A1/en thoroughly? It states under 0065:
"In essence, flow-limitation is a condition in a collapsed tube conveying a flow where (given that the upstream pressure is held constant) the flow is no longer increased by decreasing the downstream pressure (i.e., an increase in the flow-driving differential pressure)"
And it's certainly not mainly about area but of shapes instead (0067):
"A feature of flow limitation is that while the downstream pressure is sufficiently low to keep the tube collapsed the flow-rate will be more or less maintained at a constant value, regardless of changes to the driving pressure. In a patient with flow-limited breathing this equates to an inspiratory waveform with a flat top (i.e., a constant inspiratory flow-rate.)"
In 0083ff a description follows how to compute from fuzzy index variables -- related solely to the waveform -- a *flow limitation measure* after de-fuzzifying. 0094 is also worthwhile reading as ist describes how "flow limitation *may be detected* [...] from the measure of respiratory flow".
And that's just the beginning. In 0267 they even calculate the Shannon entropy of the normalized snore vector ...
Mike
RE: Why don't these flow rate curves count as flow-limited? - 2SleepBetta - 07-18-2021
(07-16-2021, 02:33 AM)multicast Wrote: Well, not "really": A real, physical flow limitation is something that happens with your body, it's not about graphs. A happening flow limitation is a non-linear response
between your breathing efforts and the amount of air you're actually breathe. That is: your belly is moving up and down but there's only few air reaching your lounges. It's not possible to detect a real FL with our machines for certain.
(07-18-2021, 02:30 AM)cathyf Wrote: Let me restate that...
"As an aside, I think I now understand what detecting a flow limit is about. It's NOT about looking at the tippy-tops of those inspiratory flow rate waves like entrails of sacrificial goats. It's all about how the area is distributed under the curves."
One member's views after seeing the contrast between the two quotations above and much else:
First, in this quotation and mode of "restating", I most lately responded elsewhere above to content of your statement, cathyf (as it is in default font before the bold below), with me using bold font, as follows: [You] "It's not about those tops that look like the letter "M"" (Quote marks added--added back?--here as I believe it is a quote for which I admitted omitting quote marks above.)
[me] True. Those are just signals, first clues, for those who have some or a lot of low level flow limitations, and they are far from the whole story for many and may become more serious and will remain when things become worse.
When a member complains of chronic, unrestful, low AHI sleep, the tips give good clues to inspiratory flow limitation in their frequency, density on the time line and/or extent of deformation. Some of the M tips, the majority of mine, even many in their isolation, are followed in the exhalation of the just-inspired amount with a FL flag--it's a Resmed FL whether on all fours or a single leg, as it considers the four factors before raising that flag. A lonely sigh among regular tips sometimes, in my case, will have that trailing FL flag, a small one. (Yeah there's a 1 to 2 two wave TV-pattern disturbance, a sudden drop of I/E, a sudden respiratory rate disturbance--IMO, a confusing FL flag-marker storm). Just nonsense? No. Change, change I harp and harp on as triggers of, or as cessations of, conditions for raising, maintaining, or ending a FL flag.
So, yes "it"--one key to a members discovery of his/her inspiratory flow limitations (IFL) can be, as you figuratively insist, about reading the "entrails", when FR shows little flattening, especially if there are few FL flags to boot.
Further, as you are aware and boiling all down--I have long and slowly stumbled along working to show that, (1), having IFL is another way of saying you have inspiratory volume (TV) losses--so Yeeeaaah! And as in My Fair Lady, Harrison's line I may butcher, "By Jove she's "got it"--cathyf's "got it", got it that 'it's" about what areas lie between the inspiratory curves and the axis below them, (2) contrary to old and newer claims and to skepticism and dismissals in the FL Expression Thread and elsewhere, a Resmed FL flag is meaningful in signaling from one to four things--TV drops, peak flattening, other mis-shaped waves, respiration irregularity, and (3), as in point (2), IFL is, though related, different from and can be more insidious IMO than what flow limitation may be indicated by a FL flag. More insidious? Because some people with high IFL, you for instance, have shown lots of symptoms in deformation of the tips and seem to have only recently, in your histrory, began to notice them at all. When? And only to dismiss them more now?
However, like trying to spot real arousals and mini-arousals, it's difficult to decide about M-tips and other +tip deformities because of cardiogenic waves. Those being most obvious at ends of exhalations, but those can also be and be misinterpreted at peak inflow. Both are points where, on one hand, there is a relatively long low in in flow rate, and, on the other hand, a short period of high flow at peak FR--can I call that an inflection point? Nah, just a local maxima; but the rate of change of inflow is very briefly near zero where c-waves tend to show. So, again, the research done after Aitokallio's 7 shapes, to distinguish 47 relevant shapes is not nothing.
Call Mr. FL or the devious Messrs, Fow Limitation ghosts if you will. But he or they leave footprints in our loss of inspiratory flow, small or large, and we call it inspiratory flow limitation or FL as the case may be.
RE: Why don't these flow rate curves count as flow-limited? - Geer1 - 07-18-2021
Your leaks were real. You can see them increase as the pressure increased and then the large leak was after your arousal when you were probably changing position etc.
"how can the machine even properly pair inspiration with expiration?"
Not sure what you mean by this. The machine does not "pair" inspiration with expiration. It supplies pressure when inward flow is detected and then stops supplying it when flow stops, when your breath gets weird like that it does struggle a bit which can be seen by your mask pressure graph.
RE: Why don't these flow rate curves count as flow-limited? - cathyf - 07-18-2021
At 1pm I'm smart enough to realize that trying to clarify something said at 3am with something said at 2am a few days later is the very definition of boneheaded.
When I look at graphs, in general, I'm looking with my math-major eyes. And when my math-major eyes are looking at some textbook example of a non-sleep-disordered flow rate graph my smart-alec math-major mouth is saying, "Hon, that's not a sine wave!" I naturally have an almost visceral relationship with graphs -- it's what the shapes feel like in my brain. Intuition is a totally different abstract notion of "feel" as opposed to what my throat is doing which I can also sometimes "feel" which is literally feeling what's happening in my own body. So when people start the whole conversation talking about sine waves, I'm scratching my math-major head trying to figure out what they are trying to get at. Because what I'm looking at doesn't feel like, look like, a sine wave.
To move to the Greek alphabet because that's got the shapes that I'm looking for, the usual discussions of flow limits always use M prominently. Shifting to Greek, everyone wants to talk about Μ (capital Mu). This is an analogy, of course. What I was trying to say (badly, because middle-of-the-night internet posts are but one step above talking in my sleep!) is that I think that what I am finally getting an intuition for, is that when looking at a flow rate graph, breaths that are not flow-limited are Λ-shaped (capital Lamda), while flow limit shows up in the graph looking like Π (capital Pi). Which is not different from what you guys have been trying to pound into my hard head all along! The whole Μ discussion is something highly rarefied going on in close-up detail in the top part of the Π, and it's not always there (because there are different sorts of flow limits).
The true moral of the story is that I've got to control the talking-out-loud-in-my-sleep part of this conversation, and as a newly-employed person after 16 months of unemployment, that's easier said than done!
RE: Why don't these flow rate curves count as flow-limited? - cathyf - 07-22-2021
Last night I tried an experiment -- sent the pressure support on the vauto to 3, to see how it compares to what the air10 with EPR at 3.
The vauto at PS of 3 is much better than the EPR of 3.
Screen Shot 2021-07-22 at 10.47.46 PM.png
[attachment=34017]
https://www.dropbox.com/s/69d6cmtnotbscq6/34017_Screen%20Shot%202021-07-22%20at%2010.47.46%20PM.png?dl=0
I was asleep during those long periods of no flow limits. At no time over the last 6-3/4 years have I ever seen a flow limit curve even close to that on the Air10 where I wasn't fully awake.
Does anybody know how the vauto uses these other parameters to knock down the flow limits? Here's my settings:
Max IPAP 25.0
Min EPAP 4.0
PS 3.0
Ti Max 2.0s
Ti Min 0.3s
Trigger Med
Cycle Med
Well, so much for my theory that if the Air10 could just be set to do EPR at 4, it would become a vauto. Looks like it's more complicated than that!
Here's a closeup of that wicked hypopnea:
[attachment=34018]
https://www.dropbox.com/s/1je0rfgtcyq5rpl/34018_Screen%20Shot%202021-07-22%20at%2011.36.35%20PM.png?dl=0
RE: Why don't these flow rate curves count as flow-limited? - Geer1 - 07-23-2021
Either medium trigger cycle triggers a bit earlier than EPR, pressure rises slightly faster or it doesn't really make sense why there would be a difference. I haven't had the opportunity to directly compare the two on same pressure/PS settings so would be curious if you see any difference when comparing mask pressure to flow rate charts between the two machines.
RE: Why don't these flow rate curves count as flow-limited? - Sleeprider - 07-23-2021
You are not the first to notice that EPR 3 and PS 3 are not the same creature; however they graphically appear the same in mask pressure (prove me wrong). I love your math major approach and appreciate someone that actually know what the 95th percentile means. We can take statistics to a certain point, and assign priority to different parameters, but in the end, we will still ask, "how do you feel"? Our objective at Apnea Board is not to get the best numbers, but to get the numbers to a point where we can get the best sleep. So we are both quantitative and qualitative in approaching therapy. What are the chances of that?
RE: Why don't these flow rate curves count as flow-limited? - cathyf - 07-23-2021
Here's last night,
https://www.dropbox.com/s/oco9df63ofmfnae/Screen%20Shot%202021-07-23%20at%209.53.39%20PM.png?dl=0
compared with a good night on EPR 3 from about 6-7 weeks ago:
https://www.dropbox.com/s/78ucovqlwh8ev54/Screen%20Shot%202021-07-23%20at%209.53.22%20PM.png?dl=0
Here they are stacked one atop the other
https://www.dropbox.com/s/n090u562ygye34j/Screen%20Shot%202021-07-23%20at%2010.09.35%20PM.png?dl=0
Is there anything that would help as a closeup?
I'm going to do another night with a PS of 3. Next logical thing to try is PS of 5 -- should I be cautious about that?
And, hey, I have to be careful and systematic about this. I'm composing a letter to my sleep doctor along the lines of I want them to lead, follow, or get out of the way. I'm anticipating all of the various arguments and trying to lay out a coherent story. The one thing that I'm worried about is if they claim that the vauto calculates flow limits differently than the Air10 and so the vauto isn't really better.
(And of course I know what the 95th percentile is. Doesn't everyone?
I've even noticed that the "max" flow limit looks more like the 99th %-tile not the max. That table is nice and neat, but for most things the Min column is useless -- maybe a 1st or 5th %-tile would be more interesting than a Min?) Here are screenshots of the first 16 days of vauto use:
https://www.dropbox.com/sh/deu5yqa19zu7wi3/AAA9X2kfkZkC57zKoaI0pNwZa?dl=0
That screenshot above from June is about the best I would ever get on the Air10 -- the vauto is just astonishingly better.
(I like doing this with dropbox. Don't have to worry about managing a 15MB attachment space, and I'm not costing apneaboard bandwidth charges...)
RE: Why don't these flow rate curves count as flow-limited? - 2SleepBetta - 07-24-2021
Gotta admire your astute analytical depth and persistence, but had to laugh recalling that I too questioned whether the VAuto simply computed things differently and that would account for some of the dramatic improvement I saw in my change (drop in FL) from Autoset to VAuto.
Can't say my old eyes discerned a whole lot looking at your best Autoset night and at your stacked comparison to the VAuto's early results. It did appear that the VAuto FL were significantly lower than for the Autoset's best night.
The larger difference, though, tends to confirm my belief, probably some of what I have read here: the Autoset tends to drop pressure too quickly and lose ground to FL, vs the VAuto as is easily seen in changes in the Mask pressure curves. I don't know how that relates, but it may, to SR's (or was it Geer1's) comment about better trigger setting by the VAuto.