Spisni study

[Wookie]

but I did have the GPSs for cars in mind, not maritime applications

I knew that. But it played perfectly into my point, and for that I will be forever grateful. You gave me an opening into the thread.

 

[Patoux01]

Gaming the software with RD2.0 compared to ZH-L-16 with GFs, what do you feel it looks like, roughly?
30/70 or something to that effect?

Well you're the one saying there's "an algorithm" behind it that is perfectly valid. I ask which is it? So far, I've heard something around 5/70 in a profile given here (was rather benign) and kevrumbo suggested something around 15/70 *on that specific dive* https://www.scubaboard.com/community/attachments/dive_to_170_profiles-png.412330/ . My guess is that it breaks down when pushed around a bit, ie it is not consistent.


So what "tested algorithm" is behind RD, that it has been able to get fudge-factored from RD1 to RD2, which both still look nothing like a buhlmann deco profile to me?

 

[Dan_P]

Well you're the one saying there's "an algorithm" behind it that is perfectly valid. I ask which is it? So far, I've heard something around 5/70 in a profile given here (was rather benign) and kevrumbo suggested something around 15/70 *on that specific dive* https://www.scubaboard.com/community/attachments/dive_to_170_profiles-png.412330/ . My guess is that it breaks down when pushed around a bit, ie it is not consistent.

So what "tested algorithm" is behind RD, that it has been able to get fudge-factored from RD1 to RD2, which both still look nothing like a buhlmann deco profile to me?

As for what's different in RD2.0 in relation to Spisni and deep stops emphasis, it has a - relatively - great reduction in deep stop emphasis by way of 1) reducing first stop depth from 75% to 66% and 2) changing the S-curve to place shallower emphasis. As you'll recall, the RD-sample group of the Spisni-study used RD1.0, which has a higher emphasis on deep stops.

But I think it's worth reiterating that it's not to be confused with an algorithm.

Just to be certain, I'm not saying RD is built to mimic one specific algorithm, but I will say that it incorporates different aspects of decompression theory - including of course the work of Buhlmann, and Wienke, etc., and I will say that I find nothing radical in utilizing the 30-minute compartment to derive a max deco-based relation to O2-time.

I don't know what you mean by the concern of inconsistency you mention, but I don't mind formulating my thoughts on it if you'll help me by narrowing down, for instance, circumstances where your concern might be relevant?

 

[Dr Simon Mitchell]

No... your version of ISS and the VPM formula are not related..... your version of ISS is one giant number .... goes against the rules and concepts of parallel mono-exponential models, duplicates data, has no concept of relative stress, un-calibrated, not verified.

I'll address this first, because it becomes relevant below. As I pointed out, it is not "my" concept; it is a very simple construct employed by the world's leading decompression modelling team to interpret the NEDU study, and it is at the core of VPM and several US Navy algorithms. There is only one "version"; it is a simple integral of supersaturation and time, and it is obvious to anyone reading this thread that time matters. It takes time for bubbles to form, so after a dive if you are supersaturated for 1 hour, that is way worse than one minute.

You have railed so strongly against this very simple fact because the integral of supersaturation and time is a metric that exposes the flaws of excessive deep stopping more clearly than any other. As for "going against the rules of mono-exponential models", this is just nonsense as Dr Doolette has explained to you before. Compartments are just mathematical constructs, not the reality of physiology, and it is perfectly valid (indeed necessary) to sum supersaturation across compartments in order to appreciate decompression status. Though it will not (unfortunately) be obvious to many readers here, this "going against the rules" notion is as stark (and scary) an illustration of your lack of understanding of decompression physiology as your recent claim that exercise and tissue perfusion does not affect decompression.

*********************************

On to the issues of this thread:

I don't have any information on O2 bio-markers causes and bio-physics interactions.... not my job.

That is not a very solid platform for opening this thread with a post that concluded:

"The elevated ppO2 extra time at 21 and 18m, causes extra off gassing and lowering of stress mid dive, thanks to higher ppO2 pressures, but it comes at the cost of elevated inflammatory markers".

In drawing this conclusion, you completely failed to explain how an air dive with low oxygen exposure but moderate decompression stress produced the same pattern of inflammatory activation as the UTD-RD profile. This observation was cited by the authors in their explanation of why they did not believe your version of events, but you simply ignored it. If you are going to analyse a scientific paper and publicly contest the conclusions of the authors, it is extremely dishonest to just ignore cogent arguments made in their published manuscript against your position.

You have repeatedly ignored my challenges to explain why you interpreted the data in a scientific paper differently to the authors, and "I don't have any information, not my job" is the best answer you can come up with.

• For the bio-marker changes, you cannot demonstrate or show any relevance to supersaturation whatsoever?
• No math or graph to show whatsoever.

The following figure kindly provided by UWSojourner shows a comparison of integral supersaturation between the two tested profiles. There are two other profiles in the diagram, which we will discuss later, but for now just focus on the UTD-RD profile and the GF 30/85 profile. In the little boxes at the top of each bar RT = run time in minutes, and the numbers represent the integral supersaturation in millibar-minutes associated with that profile.

The first thing to note is that Ross's claim that "the UTD-RD plan has less gas pressure (supersaturation) stress levels over the whole dive" is wrong (or misleading at the very least). The surface supersaturation (when bubbles are forming) in the UTD-RD profile is about 10% higher. This seems a subtle difference, but it is in the right direction for explaining the result. A similar comparison in the NEDU study showed about an 18% difference between the two profiles, and that was sufficient to produce higher bubble counts and a significant difference in DCS rates. 10% is a smaller difference and it is not surprising that we consequently see a less dramatic result. But remember, these supersaturation calculations are just numbers, not physiology. The true differences could be more (or less) but with the comments about oxygen made previously in mind, decompression stress remains the most plausible explanation.

Now look at the other two bars (the VPM 6.2 and GF 50/54 ones). Those show the supersaturation stress if you use VPM on high enough conservatism or a GF profile with parameters chosen to produce the same decompression time as the UTD-RD. Dan-P, you might want to pay close attention to this.

We have always said that having two approaches with different total decompression times was not the best design for a study attempting to see whether distribution of stops shallow or deep makes a difference. As a corollary, we have also said that 30/85 was not the optimal comparator for UTD-RD in this study (mainly be cause it over-emphasises deep stops too). The diagram provides a stark demonstration of what happens if you choose decompression profiles of the same length but with less emphasis on deep stops. Even VPM does substantially better, though Ross cannot claim any victory here because he has opined (incorrectly I might add) in the past that VPM used on high conservatism (eg VPM-B 6.2 as illustrated here) is outside the model parameters. But look what happens if you choose a GF decompression designed to be the same length as the UTD-RD profile. There is a dramatic reduction in decompression stress.

I would just like to thank UWSojourner for going to the trouble of producing this figure (and others which I have not used here).

Simon M

 

[Storker]

most research is based on a hunch, aka theory

In in science, a hunch is in no way a theory. At best, a "hunch" is a hypothesis. Only when several experiments have failed to falsify the hypothesis, and it has predictive powers, it can be called a theory. As in "theory of evolution", "theory of gravity", or "theory of relativity". In science, a theory is strong. It's something that's really useful.

Hypothetico-deductive model - Wikipedia
Scientific theory - Wikipedia

Sorry, but this is a pet peeve of mine. When a scientist refers to a theory, it's something which is very different from when your favorite crime series character says 'I have a theory...'. And since it's a very common misconception which is very often exploited by anti-science activists, hearing people repeat it raises my hackles.

 

[rossh]

VPM-B 6.2 as illustrated here

Simon M

VPM-B 6.2 ??? There is no such thing. It's a made up fake profile.... which should tell you something about the relevance of these charts.

For people who don't know.... this is how Simon' and Kevin's ISS version works... They take all the time supersaturation occurs, and add all those minutes together. Then they add all 16 cells together, to arrive at one big giant number. So there is overlapping cells that duplicate data, (not allowed in a parallel model design).

But worse still is it has no concept of stress or risk - everything is weighted the same. So 10 minutes at 4 ATA (which will kill you), is the same as 100 minutes at 0.4 ATA, which is harmless.

Junk science Simon... not verified, un-calibrated and useless for measuring stress.. And its biased to the shallow profile every time. But it does look pretty.

No, the NEDU does not use such a nonsense like your version... It does not make those mistakes above, and neither does VPM.


And getting back to the topic..... where in those bar charts, is the elevated bio-markers being generated?

.

 

[Storker]

 

[Wookie]

Well then.... I guess we see who looks silly now....

 

[Dr Simon Mitchell]

VPM-B 6.2 ??? There is no such thing. It's a made up fake profile.... which should tell you something about the people who create and promote these charts.

As has been pointed out to you many times (by multiple scientists), +any number up to at least 7 is well within the parameters of the original VPM model. If you want to argue about this, I will link back to the original discussions on RBW. You appear not to understand the model that you write software for.

For people who don't know.... this is how Simon' and Kevin's version works... They take all the time supersaturation occurs, and add all those minutes together. Then they add all 16 cells together, to arrive at one big giant number. So there is overlapping cells that duplicate data, (not allowed in a parallel model design). It has no concept of stress or risk - everything is weighted the same. So 10 minutes at 4 ATA (which will kill you), is the same as 100 minutes at 0.4ATA, which is harmless.

Except that in UWSojourner's analysis all 4 decompressions start from the same common gas loading point (the same 50m dive for the same bottom time) so the comparisons are completely valid.

Junk science Simon... but it does look pretty.

No the NEDU does not use such a nonsense like your version in its work... It does not make those mistakes above, and neither does VPM.

From: the NEDU Deep stops study report:

See the bit where it says "PDCS is a function of the sum of similar risk functions from all relevant compartments" which you say is "not allowed" in a parallel compartment design.

Not only did they use integral supersaturation, they also summed it across compartments in both the deep stops report analysis, and in their decompression algorithms (all of which you have been told before).

Simon M

 

[rossh]

As has been pointed out to you many times (by multiple scientists), +any number up to at least 7 is well within the parameters of the original VPM model. If you want to argue about this, I will link back to the original discussions on RBW. You appear not to understand the model that you write software for.
Simon M

There is no such thing as +7... its all made up fake nonsense data. Some arbitrary limiting number does not give you free license to make stuff up.

Except that in UWSojourner's analysis all 4 decompressions start from the same common gas loading point (the same 50m dive for the same bottom time) so the comparisons are completely valid. Simon M

... and your version of ISS has no relevance to stress, or any way to extract that information. ... making the whole exercise a waste of time.


The lowest stress dive, is the one that is the lowest supersaturation in the dive. That's indisputable. Which means it scores a bigger number on the Simon/Kevin ISS. Kinda works backwards.

See the bit where it says "PDCS is a function of the sum of similar risk functions from all relevant compartments"

Not only did they use it in the deep stops report analysis, they use it in their decompression algorithms (as you have been told before).

Simon M

No they don't. See the word "relevant" in the description? The NEDU are smart enough to keep the cells sparsely spaced, so they don't overlap the same data.

But in your Simon/Kevin version of ISS, it overlaps everything, and keeps compounding together, over and over duplicating data... which is why you get such a big useless number.

.