Forum: help

RE: Help to obtain Power numeric results [ Reply ]
By: Abraham Otero on 2020-06-30 08:18

Energy in signal processing in general (and HRV analysis in particular) is not, strictly speaking, the same as the notion of energy in physics. But they are analogous and related concepts.

If we have a signal x(t) with unit U, the units of measure for the signal energy are U^2·seconds, which is not dimensionally correct for physical energy. To make things even more confusing, the units of spectral energy (which is really what we are talking about here) would be U^2·seconds^2, which are different units of the energy of the original time series. However, the energy of the original time series and the spectral energy are equal in magnitude.

In the case of HRV, if we want to consider that the units U are seconds (distances between RR waves) strictly speaking the units of the spectral power would be seconds^2·seconds^2= seconds^4. If we were working in milliseconds they would be ms^4.

In practice, I wouldn't worry too much about all this. Typically what we are interested in studying in HRV are statistically significant changes (potential biomarkers) between two or more populations in some index of the HRV (spectral power in a band in this case), and ideally, to find a pathophysiological explanation that allows us to interpret that change. The energy of all subjects is always in the same units, so it can be directly compared. Understanding exactly the units or methodology used to compute the spectral power calculation (of the HRV or any other time series) requires understanding relatively complex math that unless you have a background in electronic engineering you will probably miss some of the details. But you definitely do hot need to know these details to use a tool like RHRV or Kubios to find biomarkers.

RE: Help to obtain Power numeric results [ Reply ]
By: Claucio Rank on 2020-06-29 19:30

[forum:47897]

Thank you Mr. Otero,

so the energy measure unit is like s^2?

The n.u parameter is a normalized data of LF and HF (like the percentages), but without the VLF band.

I used to do the HRV analysis with Kubios, I´m new on RHRV and I intend to use it in my laboratory experiments since it´s a more powerful and practical way to do it, the only difficult that I´m experiencing is this little detail because sometimes I will need to describe this data.

Regards.

RE: Help to obtain Power numeric results [ Reply ]
By: Abraham Otero on 2020-06-29 12:23

[forum:47888]

The calculations of the power that RHRV performs would be equivalent to what Kubios calls ms^2, although RHRV works in s (seconds). The percentages are easy to calculate, as it is simply what percentage of the energy falls within each band. What I don't know is what exactly n.u. is...

Are you trying to compare the results of RHRV with Kubios? This is not easy since there are a large number of parameters that are used in the calculations, many of which are not reflected in the user interface, and since they cannot see the source code, it is not known exactly what value they have used.

Help to obtain Power numeric results [ Reply ] By: Claucio Rank on 2020-06-25 19:41	[forum:47886] Captura de tela de 2020-06-25 16-29-46.png (27) downloads
Hello, how could I, obtain the power band frequency domain results in ms², n.u and % like in Kubios(anexed image). Regards, Claucio.