Hello David,
After applying the dfa A1 metric one potential it has if the correct protocol is done is the extraction of thresholds.
I have the formulas that have been created for estimation and then to be able to analyse correctly. Would it be possible to implement it?
HRV data is currently collected by intervals but is not being processed. (DFA A1 is based on the HRV data)
current DFA a1 data is being processes “AS IS” from usage of the alphaHRV data field from Garmin Connect and will be made visible on the activities (if it contains the data)
David has indicated that processing of the HRV data for DFA A1 is in his radar.
Hello,
I know how AlphaHRV works and how it calculates through RR intervals, in fact I am an IQ collaborator, I have more than 100 tests performed, “standardized” protocols, machine learning systems.
With some very simple formulas generated with wko it is very easy to estimate the thresholds, that’s why my offer to David to be able to implement it without having to do a lot of research. We think that when dfa enters at 0.75 we have the threshold and it is not like that, there are many times that you can enter at 0.75 and leave again, you have to analyse the last entry which is when it does not leave 0.75 again, the same for 0.50. You have to use a protocol that is in all the papers of Bruce Rogers and soon in that of a Spanish research team in which I collaborate.
My offer to David still stands to facilitate his work based on my experience.
Tx. I will definitely be in contact when I get onto this. Figuring out thresholds from DFA a1 data would be neat.
David,
I will be happy to help you with this. When you need to implement it I will give you an explanation of everything we have done and the formulas we use in wko5 to calculate the thresholds.
Correct if I’m wrong but aren’t you and @Inigo_Tolosa THE ones of alphaHRV IQ app?
If so, very humble of your part.
Regards
Hello
I’m Íñigo’s colleague, I don’t mind offering my help to a platform like intervals, we don’t get economic benefit from all this, intervals has a lot of future and the more functionalities it has the better. For me it is an honour
Is there a reason the AlphaHRV app doesn’t work on older generation devices, eg. FR 735XT?
I think that is a problem with memory
Memory and processing problems
Thanks, it’s a pity as the watch can provide such a lot already.
Sorry top resurrect such an old thread, but it is such an interesting topic.
@Luisma_Gallego_Soy_P what are your thoughts on the recent paper in which Bruce Rogers talk about setting a custom Alpha-1 point to better reflect HRVT1 (rather than 0.75): “the novel finding presented here indicates that this target needs to be derived from the maximum DFA a1 seen early in an individual’s ramp incremental.” 1
If this is validated, it seem to potentially also be doable in the AlphaHRV; I’d imagine the app can do the calculation automatically?
E.g. identify highest a1 point during the first zero to five minutes window (or zero to ten minutes window; the paper is quite vague), calculate the mean followed by a custom message like the readiness blob: your custom HRVT1 a1 is: xx ?
In my opinion, that customized alpha-1 value is going to change every day, so looking for a personalized value for each athlete and then relying on that value honestly doesn’t make sense to me. I don’t know how Rogers continues with fixed values (I haven’t read the study, just the article on Rogers’ website). Personally, I use a completely different approach. My approach can detect hrvt1 in any training session where there is really intensity above the first threshold. I will review Rogers’ article and see what I can create. Rogers’ approach talks about an incremental test; I don’t do incremental tests to detect it.