ResMed Adapt SV vs. Respironics Auto SV

Speaking of syllogisms, they are a huge caveat in general when it comes to analyzing any complex system behavior. They show up in this thread in spades.

SAG wrote:And if that was it's function, then it would also be on Malibu, which, it is claimed, has the same breath support as AdaptSV.

Here's one such example. These two machines do not share the same breath support. Resmed merely claims" "The smooth pressure waveform, taken from the adaptive servo-algorithm of the VPAP Adapt SV, mimics normal respiration so that breathing feels more natural". Specifically, Resmed is referring to that comfortable curved pressure rise and saw-toothed decay that these two machines have in common.

But that's where the breath-support commonalities end. The ASV is targeted for (primarily PS-based) treatment of central-dysregulation variants and the Malibu is targeted for (primarily static-pressure based) treatment of OSA variants. The ASV thus accomplishes its breath support by holding a static-pressure bias constant while fluctuating PS. The Malibu accomplishes its breath support by holding PS constant while very gradually varying its static-pressure bias.

Most importantly, the ASV algorithm relies on instantaneous flow-point samples to support instantaneous PS-based impeller-rotations. By contrast the Malibu quite happily works with a less-precise derived approximation of tidal volume toward very gradually changing its static-pressure bias.

These two machines really do not "share the same breath support" across the board by any stretch. The ASV is all about instantaneous detection and instantaneous response. That's what drives the ASV's need for signal sensitivity toward picking off all those transient-based sinusoidal reference points. That same functional design requirement does not exist in the Malibu, which suffices with tidal volume approximation quite nicely. These machines are not only apples and oranges regarding system output ("breath support") but they are apples and oranges regarding system input (analog sampling requirements). This lack of requirement for signal-processing transient sensitivity is precisely why the Malibu has no proximal sensor line.

Unlike the Malibu the ASV machine strings together many instantaneous "vertical slices" of flow-curve sampling and equally instantaneous impeller response. That heightened need for instantaneousness is precisely what drives the need for heightened sensitivity to the reference-point dynamics and transients that are embedded in a proximal-based airway signal source.

SAG wrote:So anyway, if what -SWS implies is correct, then it would seem that jamming the Paw signal would really create some havoc now, wouldn't it?

Banned wrote: SAG must be correct, disconnecting the Paw only leads to an annoying alarm and 'Low Pressure - Check Circuit' message.

Here's a potential syllogistic caveat IMO. How's this for comprehensive testing methodology? Disconnect or jam the proximal signal, observe the machines' response, then make a rational call as to whether the system really needs that proximal sensor line.

(Now why does this improvised methodology also remind me of that time SAG introduced "real world variables" to show the world how dangerous Resmed EPR is? Crikey, mate. No wonder some people understandably wonder to what extent CPAP manufacturers might be here "working" the message boards ).

In the case of Banned's experiment all we see is that the system raises a white flag that essentially says: "Hey I have a serious problem with my input signal". However, the system output by the way of pressure response appeared sane. And that sane appearance doesn't mean that the need for proximal signal sensitivity is obviated. Rather, the "somewhat sane" sustained output of this system speaks of a designed functional feature known as "fault tolerance". In the case of Banned's ASV experiment fault tolerance sustains recently-derived system output while generating an alarm for the user.

My problem with many of these ad-hoc experiments is that they are very understandably myopic. So far none of the ad-hoc cursory experiments I have seen demonstrated on this message board would have sufficed as conclusive methodology when I worked in R&D for ten years. I don't mean that statement in any sort of harsh way. I simply mean it as a well-intended reality check toward some of the scientific conclusions I have seen cleverly but incorrectly derived on this message board.

-SWS wrote:
<snip>

But that's where the breath-support commonalities end. The ASV is targeted for (primarily PS-based) treatment of CSA variants and the Malibu is targeted for (primarily static-pressure based) treatment of OSA variants. The ASV thus accomplishes its breath support by holding a static-pressure bias constant while fluctuating PS. The Malibu accomplishes its breath support by holding PS constant while very gradually varying its static-pressure bias.

<snip>

On this point we see a similar difference between the Bipap AUTO and the Bipap SV. I had in the past speculated that the Bipap SV was an amalgam of the Bipap Auto algorithm with timed mode from the Bipap S/T & placed in the Biap S/T case.

But, it is now clear from all the discoveries (for me) of the past month including the Resp Q&A that the Bipap Auto varies the static pressure independently for both ipap & epap (auto titration) whereas the Bipap SV has *no* auto titrating algorithm in it. It holds epap static but will vary ipap based on flow sensing (not OSA event sensing) and mixed in with breathing rate and depending on what mode was set.

Things get confusing when the names of these 2 machines both contain the word Auto, but the auto means 2 quite different things. For the Bipap Auto it means Auto titration (of epap and ipap) but for the Bipap AutoSV it means Auto rate & flow adjustment.

Both are Bipaps though.

DSM

-SWS wrote:
<snip>


These two machines really do not "share the same breath support" across the board by any stretch. The ASV is all about instantaneous detection and instantaneous response. That's what drives the ASV's need for signal sensitivity toward picking off all those transient-based sinusoidal reference points. That same functional design requirement does not exist in the Malibu, which suffices with tidal volume approximation quite nicely. These machines are not only apples and oranges regarding system output ("breath support") but they are apples and oranges regarding system input (analog sampling requirements). This lack of requirement for signal-processing transient sensitivity is precisely why the Malibu has no proximal sensor line.

<snip>

SWS,

Your points esp this one seem self-evident.

Recent posts (esp from you) have really brought home to me the dramatic difference in the nature of the SV machine vs an Auto titrating machine (be it AutoCpap or AutoBipap).

I am reminded of the collective myopia we all had about Autos when they were such a hot topic in 2004/2005. There were respected people here actively stating that the Auto would detect a block and clear it. The way most people took this (from the way it was put) was to conclude that Autos cleared blocks in real time.

When we got into the reality of the algorithms and how long it takes an Auto to raise its pressure even by 1 CMS, it became very clear that Autos can only *slowly* adjust pressure in response to active events and to some pre-cursor input and that the logic of the Auto is that by raising pressure reactively, the next event may be pre-empted but if no other active events or monitored pre-cursor input is detected, for a set period, pressure can be lowered again.

Autos attempt to clear 'patterns' of obstructions.

The Adapt SV's instantaneous reaction to events is in one sense what we used to think Autos did.

So a really big difference between the SV and an Auto is that SV machines can switch from passive to active ventilation instantaneously whereas Autos offer no such capability.

DSM

#2. RE Autos, I also recall that some published papers also said out that when an obstructive apnea was actually in progress, increasing pressure too quickly or too much during the apnea could be dangerous. Also that Autos mostly guessed at if the block was obstructive or a central & that raising pressure blindly when a central was in progress was also dangerous which was why some machines would not exceed 3 CMS in a single set of rises. But, as we know, SV machines do clear centrals with increased pressure & quickly & lots of it .

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): auto

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): auto

Agreed, Doug. The auto machines that are aimed at treating obstructions primarily strive to preempt those obstructions. When they gradually raise static pressure, they are inherently gradual by system design (because of epidemiology-based precaution). The gradual static-pressure increases were never because impeller technology or general processing was incapable of quicker rotations.

The ASV's need for rotational speed and efficiency is because of an extremely heightened need for instantaneousness---instantaneousness certainly on the output side regarding the precise number of impeller rotations that must be delivered now. They must be delivered "now and with precision" to instantaneously serve those vertical and rather narrow slices of PS-based and even F-based discrete targets (instantaneously targeting pressure support and respiration rate respectively).

Endeavoring to very quickly pressure-target a collection of phase-based reference points undoubtedly contributes to that need for such extreme instantaneousness regarding both sampling and response. However, a part of the ASV's heightened demand for such extreme instantaneousness undoubtedly has to do with it's highly unique task of endeavoring to instantaneously adjust patient variable "F" (respiration rate) by instantaneously adjusting machine-variable "F". It accomplishes this latter task by pinning up those instantaneous PS targets (that are reference-point based) in an either advanced or delayed delivery manner of discrete sequencing.

From what I have seen, the air valve in the Bipaps can also adjust airflow in millisecs and thus is a particular way of providing rapid response if called on. The air valve appears to be very very precise.

It has always struck me that the current Bipaps, if they wanted to, could do exactly what the Vpap SV achieves but in order to safely do it they would, IMHO, need proximal sensing. If we agree that the air valve is equal in responsiveness (maybe better) than the twin impeller motor in the Adapt SV then it is left to the controlling algorithms and sensors.

I suspect though, that the real benefit of the twin impeller motor design is lower cost to manufacture.
#4 Actually, it is plainly obvious that one self contained blower unit (adapt) is going to cost a lot less and take up less space than an air valve *and* a blower motor (Bipap).

If ever Respironics fit a proximal sensor line to the Bipaps, I'll know they have decided to address exactly the same specialist segment as the Adapt SV. The Adapt SV targets CSR, CA, & periodic breathing.

As I understand it the Bipap AutoSV appears to address a variant of the SDB market covering complex apnea and periodic breathing. This is evident because of the ability to run it as a straight BiLevel as well as in its version of SV mode.

DSM

#3 corrected what the SVs were designed to address.

#5 It has just occurred to me that the M series machines (Bipap models) don't have an 'Airvalve' but use a single blower. The Bipap AutoSV does have the traditional Respironics airvalve.

The airvalve is a device for controlling the flow of air and it can very rapidly adjust how much air exits port 1 of the valve & redirects the remaining air out port 2 (which is usually fed back into the input of the blower fan which itself feeds air to the airvalve). Port 1 is usually connected to the air hose going to the user.
_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): respironics, bipap

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): respironics, bipap

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): respironics, bipap

SAG wrote:Coming up next: "Sweetheart, I set the alarm clock to wake you up every 15 minutes. Could you look over and write down all the ResControl parameters?"

SAG

This is SAGWife. SAG will not be posting this morning because he is busy giving birth to a ResContol II. However, he asked me to deliver this message, given to me when I asked "Any last requests before I...":

SAG wrote:Great story, -SWS!

Didn't find any support to show that the Paw line does what you say it does, did you?

Owww! Ohhh! Owwww!!

SAGWife, warning him that he better stop his %&#$% experiments.

Didn't find any compelling support at all now that you mention it, SAGfamily.

Just plenty of Resmed literature claiming peak sinusoidal reference points are used in their instantaneous calculations... coupled with SAG's own measurements showing that same sinusoidal peak completely truncated without use of a proximal sensor line. Nothing supportive found at my end so far. How about you?



Coming Up Next: More scientific tail-chasing and entertaining sophistry from the entire SAG family!









(as PAPA-SAG tries diligently to drum up a methodology to properly factor "fault tolerance" routines in the black-box output of a complex system).

A further comment on SAG's patient & machine waveforms charts.

SAG, surely the waveform at the machine end will *never* be created to look the same as at the patient end even if there were no real patient there but a bladder (say a balloon).

The compressibility of air and the movement of air through the tube, are surely going to reshape (soften) the waveform to look more closely like what you are showing as 'patient effort'. My guess is that replacing the patient with a suitable balloon, will produce the same 'patient effort' waveform.

Restated, the point I am making is that the machine end waveform as generated based on air line pressure sensing, flow rate and proximal input data, is going to be further modified by the inherent characteristics of the hose & the receiver - and that is exactly what your 'patient effort' charts shows.

DSM

ResControl II

I just did a search on this and discovered something very interesting.

It seems that both the following devices are the same basic device ...

http://www.resmed.com/en-us/products/cl ... u=products

http://www.fda.gov/cdrh/pdf6/K063501.pdf

#3 http://www.fda.gov/cdrh/pdf4/k040944.pdf

The Knight-Control II application date April 6th 2007 basically states it is the ResControl II with a different name. It is listed as controlling PB machines incl my PB330. It seems it offers remote control as well as data recording. So it seems that the serial interface on the PB3xx & PB4xx machines may be similar to the Resmed S7, S8 and Vpap II/III machines serial interface probably including the data protocol. Hmm, wonder id ResScan will recognise the PB330 (will try but I suspect not as it probably only recognises machines with pre-set Remed model types).

So it is possible (probable) that the Knight Control II is programmed to only recognize a particular set of PB models.

Being the owner of a PB330 that I can't get data off, makes this interesting to me. Here in Australia the 7-channel ResControlII is priced at approx $AUD 1800.


DSM

#2 SAG - If I buy my wife a KnightControl II, that will trump your ray-o-vac torch & note pad approach. Can just see wife with her remote sitting up in bed at 2 am taking readings & inflating/deflating me as she sees fit

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): resmed

_________________

CPAPopedia Keywords Contained In This Post (Click For Definition): resmed

-SWS wrote:

SAG wrote:And if that was it's function, then it would also be on Malibu, which, it is claimed, has the same breath support as AdaptSV.

Here's one such example. These two machines do not share the same breath support. Resmed merely claims" "The smooth pressure waveform, taken from the adaptive servo-algorithm of the VPAP Adapt SV, mimics normal respiration so that breathing feels more natural". Specifically, Resmed is referring to that comfortable curved pressure rise and saw-toothed decay that these two machines have in common.

But that's where the breath-support commonalities end. The ASV is targeted for (primarily PS-based) treatment of central-dysregulation variants and the Malibu is targeted for (primarily static-pressure based) treatment of OSA variants. The ASV thus accomplishes its breath support by holding a static-pressure bias constant while fluctuating PS. The Malibu accomplishes its breath support by holding PS constant while very gradually varying its static-pressure bias.

Most importantly, the ASV algorithm relies on instantaneous flow-point samples to support instantaneous PS-based impeller-rotations. By contrast the Malibu quite happily works with a less-precise derived approximation of tidal volume toward very gradually changing its static-pressure bias.

These two machines really do not "share the same breath support" across the board by any stretch. The ASV is all about instantaneous detection and instantaneous response. That's what drives the ASV's need for signal sensitivity toward picking off all those transient-based sinusoidal reference points. That same functional design requirement does not exist in the Malibu, which suffices with tidal volume approximation quite nicely. These machines are not only apples and oranges regarding system output ("breath support") but they are apples and oranges regarding system input (analog sampling requirements). This lack of requirement for signal-processing transient sensitivity is precisely why the Malibu has no proximal sensor line.

Unlike the Malibu the ASV machine strings together many instantaneous "vertical slices" of flow-curve sampling and equally instantaneous impeller response. That heightened need for instantaneousness is precisely what drives the need for heightened sensitivity to the reference-point dynamics and transients that are embedded in a proximal-based airway signal source.

SAG wrote:However, let us not confuse breath support (patient is still generating effort) with breath generation once the AdaptSV starts sending in controlled breaths to overcome a central apnea. At that point, from patient perspective, the system changes from a largely negative pressure to a positive pressure one, and waveform duplication becomes meaningless.

-SWS wrote:

SAG wrote:So anyway, if what -SWS implies is correct, then it would seem that jamming the Paw signal would really create some havoc now, wouldn't it?

Banned wrote: SAG must be correct, disconnecting the Paw only leads to an annoying alarm and 'Low Pressure - Check Circuit' message.

Here's a potential syllogistic caveat IMO. How's this for comprehensive testing methodology? Disconnect or jam the proximal signal, observe the machines' response, then make a rational call as to whether the system really needs that proximal sensor line.

(Now why does this improvised methodology also remind me of that time SAG introduced "real world variables" to show the world how dangerous Resmed EPR is? Crikey, mate. No wonder some people understandably wonder to what extent CPAP manufacturers might be here "working" the message boards ).

In the case of Banned's experiment all we see is that the system raises a white flag that essentially says: "Hey I have a serious problem with my input signal". However, the system output by the way of pressure response appeared sane. And that sane appearance doesn't mean that the need for proximal signal sensitivity is obviated. Rather, the "somewhat sane" sustained output of this system speaks of a designed functional feature known as "fault tolerance". In the case of Banned's ASV experiment fault tolerance sustains recently-derived system output while generating an alarm for the user.

My problem with many of these ad-hoc experiments is that they are very understandably myopic. So far none of the ad-hoc cursory experiments I have seen demonstrated on this message board would have sufficed as conclusive methodology when I worked in R&D for ten years. I don't mean that statement in any sort of harsh way. I simply mean it as a well-intended reality check toward some of the scientific conclusions I have seen cleverly but incorrectly derived on this message board.

Well, we're about 11 posts from concluding anything, if, in fact, we are able to conclude anything at all (and unless Michael Berthon-Jones starts to post, we're probably not going to, either). Those observations should spawn thought, Mr. R&D. "Concluding" that there absolutely is a distal pressure measurement and that it's contribution is significant. It may measure everything or it may work in conjunction with the proximal pressure line. Clearly, there is also monitoring performed exclusively by the proximal line.

ALARM SPECIFICATIONS
• HIGH MAXIMUM MASK PRESSURE ALARM
Stops air delivery at 25 cm H2O >700 msec.
• HIGH MEAN MASK PRESSURE ALARM
Stops air delivery when mean mask pressure exceeds 15 cm H2O (averaged over approx 1 minute).
• LOW MASK PRESSURE ALARM
Mask pressure is less than 3 cm H2O for longer than 5 seconds.
• INSUFFICIENT PRESSURE SUPPORT ALARM
Is triggered if for 3 breaths in a row the maximum mask pressure achieved at inspiratory to expiratory transition is either <87.5% of the absolute mask pressure targeted at that point, or < absolute mask pressure targeted minus 1 cm H2O, whichever is the lower.

Pprox may participate in Learn Circuit to calculate Resistance (Pmachine - Pprox / Flow = Resistance) but I don't know if it really creates an unique correction factor or if it's simply looking for things that will mess up an existing correction factor (nasal pillows, clogged bacteria filter, etc.).

-SWS wrote:Now here's a nice chart recently posted by SAG that shows significant degradation of a flow signal measured at the machine (flow2) compared with that same patient flow signal more accurately measured at the mask (flow1):

Not necessarily absolutely identical. SAG did say he was using pressure transducers, and the way that ResMed displays their waveform (contrived or not) says it's a pneumotach. Which suggests that there are three monitoring devices, because (myopic ad-hoc experiment censored).

dsm wrote:I Can just see wife with her remote sitting up in bed at 2 am taking readings & inflating/deflating me as she sees fit

That's disgusting.

Speaking of which, better hose off the ResControl II. And it’s a girl, thank you very much.

SAG

PS - Yes, I know, saying "an unique" is like saying "an harmonica."

StillAnotherGuest wrote:

-SWS wrote:Now here's a nice chart recently posted by SAG that shows significant degradation of a flow signal measured at the machine (flow2) compared with that same patient flow signal more accurately measured at the mask (flow1):

Not necessarily absolutely identical. SAG did say he was using pressure transducers, and the way that ResMed displays their waveform (contrived or not) says it's a pneumotach. Which suggests that there are three monitoring devices, because (myopic ad-hoc experiment censored).

Of course, a pneumotach flow sensor is really a series of extremely delicate sequential pressure transducers, all wrapped up in one neat little signal-processing package. Flow is thus derived from sequential pressure readings, certainly in this and arguably in many (perhaps most) signal-processing cases in modern medicine.

And rhetorically, what happens when you signal-process a sequence of (pressure-reading based) zero-slope inspiratory curve segments? You still end up with zero slope for the resulting flow segment (as long as flow was derived as a function of sequential pressure and time). If Resmed truly wants to sample and process multiple reference points sitting on the top of a sinusoidal inspiratory peak, it makes no sense to read that sinusoidal signal way down at the machine (versus sampling with heightened proximal sensitivity at the mask). How is that added sensitivity at the mask mainly achieved? The proximal sensor line side steps the machine delivery path itself (thereby avoiding a significant portion of path-related signal-skewing affects that you personally measured at the machine very nicely as Flow2). Diameter reduction itself helps with signal sensitivity as well.

But you mentioned how important the proximal sensor line is toward calculating leaks. I absolutely agree there. Specifically, what Resmed is doing with their leak calculations at the mask has to do with picking off the stray frequency and transient signal components that are characteristic of breech-related spurious conductances. That's another case of wanting to get a narrow-diameter proximal sensor line at the mask, toward increasing analog sampling sensitivity of the dynamics/transients embedded in the flow signal stream. As a related side note, Quattro's ASV physical design optimizations probably have to do with removing resident transient and frequency signal components that just so happen to be common with breech-related spurious conductances (thus making leak-based signal differentiation much easier for the ASV algorithm when using Quattro).

Honestly, SAG, it's hard saying which of those two objectives served as the single most compelling ASV design reason to get a proximal sensor lead placed up at the mask. I only took the opposite side of that debate because we both seem to have some pretty outrageous fun when we explore analytical topics that way. However, I personally think those two very compelling reasons to have a proximal sensor line at the mask genuinely happen to travel hand-in-hand, because of the "temporally short-based" objectives and risks inherent in this type of fuzzy-based PAV algorithm.

By that I mean that the very discrete and thus highly instantaneous nature of a fuzzy-logic-driven PAV algorithm accounts for both: 1) a compelling reason to sample, process, and target multiple sinusoidal reference points, and 2) a temporally short-sighted fuzzy-logic tendency for pressure runaway (thus making increased sensitivity toward transient-based leak signal differentiation at the mask an absolute must). Here we have some extremely short-termed temporally-based objectives and risks, making a proximal sensor line at the mask a very compelling design decision (compelling toward increased dynamic and transient signal sensitivity). A heightened instantaneousness regarding functionality truly is the basis for many or most of the salient benefits and technical risks inherent in this ASV design in my opinion.

As a side note there's no runaway-inclined fuzzy PAV algorithm looming inside that OSA-targeted Malibu machine (which simply adjusts the static-pressure bias while holding PS nice and constant). Nor, for that matter, do those same fuzzy-based "temporally short-targeted" technical objectives and risks occur in the peak-flow-based Respironics SV machine (which thus also understandably does not need a proximal sensor line toward an instantaneous focus on transients and dynamics embedded in the flow signal).

At least that's my explanatory "story" for today, SAG. Carry on as you may and congratulations on that ResControl II!





Philosophical Thought for the Day: It's better to make our ideas fit science than to make science somehow fit our ideas.

Hey Snooze,

I must have missed the outcome of your sleep study. Did the doc scribe an ASV?

Banned

SAG wrote:However, let us not confuse breath support (patient is still generating effort) with breath generation once the AdaptSV starts sending in controlled breaths to overcome a central apnea. At that point, from patient perspective, the system changes from a largely negative pressure to a positive pressure one, and waveform duplication becomes meaningless.

Regarding machine-based decision making: Forget how the patient receives or "biologically processes" respiratory support (regarding whether they are on positive or negative support). The issue regarding machine functionality is truly comprised of what the machine detects as well as how it signal-processes and decision-renders during those two treatment scenarios.

What designer in their right mind would be algorithmically bound to needlessly duplicate a meaningless succession of machine-generated waveforms? The machine knows precisely when it is actively ventilating and when it is passively ventilating. Of course, remove that proximal sensor line, as you did in your experiment, and the machine has a truncated waveform to work with that looks the same in either treatment scenario. Did I mention yet that sinusoidal-based signal sensing at the mask offers fuzzy-based PAV some very compelling reasons to place a proximal sensor line in the mask?

SAG wrote:Well, we're about 11 posts from concluding anything, if, in fact, we are able to conclude anything at all..

We? Who's we? When exactly may we conclude? Thank you for being our group-think leader---especially when it comes to science lessons.

SAG wrote:Those observations should spawn thought, Mr. R&D. "Concluding" that there absolutely is a distal pressure measurement and that it's contribution is significant...

Pprox may participate in Learn Circuit to calculate Resistance (Pmachine - Pprox / Flow = Resistance)

To correctly say that distal-pressure factorization is significant does not mean that pressure must be dual-sensor measured (once at at each end) for either treatment or Learn Circuit, SAG.

Patent wording is tricky. The above patent-description wording also alludes to multiple implementations that are being legally protected. And a patent description isn't really a low-level technical requirements document or even a low-level design disclosure. The word "may" in that patent description implies that at least one viable alternate implementation exits for calculation of resistance. But Let's express that same equation in different terms:

(Pressure Loss) / Flow =Resistance

I wonder if that problematic albeit highly basic "Pressure Loss" feline hasn't been successfully skinned in multiple ways over the years. It's just a fixed highly-basic correction factor. It's ultimately used toward pressure-delivery calculations and doesn't really lend leak-based signal differentiation (mathematical filtering) a compelling advantage to make or break the proximal-line design decision (based on that fixed resistance value alone). But you "may" use Pprox if your final design implementation happens to have that sensor measurement handy. If you already have proximal measurement for those other two highly compelling reasons I mentioned earlier, then why not take advantage of it toward Learn Circuit? I know I sure would... because I "may".

SAG wrote:Well, we're about 11 posts from concluding anything, if, in fact, we are able to conclude anything at all...

I was thinking the same thing about the manufacture of tennis balls. All obvious physics observations aside, we're about 11 posts away from being able to conclude (or likely reach an impasse) regarding why they always make them so round and bouncy. All obvious physics observations aside, some trade secrets are meant to be just that!






Not to lose track of Banned's recent question for Snooze_Blues:

Banned wrote:Hey Snooze,

I must have missed the outcome of your sleep study. Did the doc scribe an ASV?

Banned

Sorry if we hijacked your thread, Snooze_Blues.

-SWS wrote:The issue regarding machine functionality is truly comprised of what the machine detects as well as how it signal-processes and decision-renders during those two treatment scenarios.

I thought the issue was determining the function of the proximal pressure line.

-SWS wrote:What designer in their right mind would be algorithmically bound to needlessly duplicate a meaningless succession of machine-generated waveforms?

If the patient is having a large percentage of central apneas (which is why people use this thing) it doesn't matter what the mental state of the designer is, that's what would happen.

-SWS wrote:Of course, remove that proximal sensor line, as you did in your experiment...

What experiment is that?

-SWS wrote:

SAG wrote:Well, we're about 11 posts from concluding anything, if, in fact, we are able to conclude anything at all..

We? Who's we? When exactly may we conclude? Thank you for being our group-think leader---especially when it comes to science lessons.

Our? Who's our?

-SWS wrote:The above patent-description wording...

That's not patent wording, that's Ohm's Law.

-SWS wrote:The word "may" in that patent description implies that at least one viable alternate implementation exits for calculation of resistance.

Right. If Pprox isn't used in the calculation then you have to use

-SWS wrote:To correctly say that distal-pressure factorization is significant does not mean that pressure must be dual-sensor measured (once at each end) for either treatment or Learn Circuit, SAG.

It does if you want to accurately account for additions/subtractions to the patient tubing, which isn't accounted for by using K1/K2. Actually measuring the difference of P1 - P2 will be able to account for variables within the circuit, such as bacteria filters, humidifiers and restrictive patient interfaces. This in turn will add the necessary precision to the volume calculation which is critical in establishing Target. Learn Circuit will either actually calculate resistance or alert the user that the circuit intended for use will not work properly (if the calculation is done internally and the K1 - K2 formula is being used).

-SWS wrote:I was thinking the same thing about the manufacture of tennis balls. All obvious physics observations aside, we're about 11 posts away from being able to conclude (or likely reach an impasse) regarding why they always make them so round and bouncy.

I think I know where the Bong went.

SAG

-SWS wrote:Right. Quibble as you may...

I wasn't quibbling!

Lemme know if you lose any of the text during the edit.

Here's a clue to beat my graph.

It's only a 30 second window.

SAG