Edit on GitHub

BRS

Several methods to estimate baroreflex sensitivity (BRS) have been defined, many of which are contained in this package. All methods require the following inputs:

  • RR: The RR Interval series in ms
  • SBP: The systolic blood pressure in mmHg

Several methods can be tweaked by using keywords, though the most common and recommended parameters are set as default.

Cardio.BRS.getBRS — Function
    getBRS(RR::Vector{<:Real}, SBP::Vector{<:Real})

Returns all BRS measures with default settings for individual methods. For more control use each method individually.

Methods

  • SME: Sequence Method
  • RMSSDR: RMSSD ratio
  • xBRS: Cross-correlation baroreflex sensitivity
  • TFBRS: Transfer Function based BRS (fft)
  • PRSABRS: Phase Rectified Signal Averaging
  • αLF: LF component of the AR based spectral decomposition
  • αHF: HF component of the AR based spectral decomposition
source

Sequence Method

Cardio.BRS.sme — Function
sme(RR::Vector{<:Real}, SBP::Vector{<:Real}; thresholdRR::Float64 = 4.0, thresholdSBP::Float64 = 1.0, seqLen::Int = 3, delay::Int = 1, minCor::Float64 = 0.8)

Calculate the Baroreflex Sensitivity (BRS) for a serious of RR intervals and respective systolic blood pressure using the Sequence Method (SME).

Keywords:

  • thresholdRR: Threshold for change in RR interval to qualify for a valid sequence (literature: 4ms)
  • thresholdSBP: Threshold for change in SBP to qualify for a valid sequence (literature: 1mmHg)
  • seqLen: Minimum length of a valid sequence (literature: 3)
  • delay: Delay between RR and SBP (literature: 1)
  • minCor: The minimal correlation between RR and SBP in a sequence to qualify as a valid sequence

Return:

Returns a SME struct for which a plotting recipe is provided, so that found sequences can be inspected by calling plot() when using Plots.jl

source
Cardio.BRS.SME — Type

Struct that stores all information related to the sequence method. The main result is stored in 'value'. It can be plotted for visual inspection.

source
result = BRS.sme(input.RR, input.SBP)
plot(result, dpi = 120)

xBRS Method

Cardio.BRS.xbrs — Function
xbrs(RR::Vector{<:Real}, SBP::Vector{<:Real};...)

Calculate the xBRS index for the assesment of baroreflex sensitivity. Based on Westerhof, B. E. et al. (2004). Time-domain cross-correlation baroreflex sensitivity: performance on the EUROBAVAR data set. Journal of hypertension, 22(7), 1371-1380

Keyword Arguments

  • minCor: Minimal significant correlation, dafaults to 0.632 (p = 0.05, two-sided for 10 s window)
  • tExcerpt: Length of the sliding window in seconds, defaults to 10 s
  • delays: Which delays, shift of the RR window shall be considered, defaults to 0:5

Return

Returns an xBRS structure, that can be be plotted for visual inspection.

source
Cardio.BRS.xBRS — Type

Struct that stores all information regarting the xBRS etimation. The final result is stored in 'xBRSg'. It can be plotted for visual inspection.

source
result = BRS.xbrs(input.RR, input.SBP)
plot(result, dpi = 120)

RMSSD Ratio

Cardio.BRS.rmssdr — Function
rmssdr(RR::Vector{<:Real}, SBP::Vector{<:Real})

Returns the RMSSD ratio := RMSSD(RR) / RMSSD(SBP)

source

Transfer Function Method

Cardio.BRS.tfbrs — Function
tfbrs(RR::Vector{<:Real}, SBP::Vector{<:Real}; ...)

Transfer function based BRS measure as defined by Robbe et al.

Keyword Arguments

  • n: length of hamming window for spectral estimation, defaults to length(RR) ÷ 10
  • minCoh: minimal valid coherence, defaults to 0.5
  • LF: The frequency range defined as low frequency, defaults to (0.04, 0.15)
source
Cardio.BRS.tfBRS — Type

Struct that stores all information regarding the tfBRS estimation. The final result is stored in 'value'. It can be plotted for visual inspection.

source
result = BRS.tfbrs(input.RR, input.SBP, n = 100)
plot(result, dpi = 120)

Spectral AR-based Method

Cardio.BRS.arbrs — Function
arbrs(RR::Vector{<:Real}, SBP::Vector{<:Real}; p::Union{Int, UnitRange{Int}} = 12:18, nfreq::Int = 256, sf::Real = 1, LF::Tuple{<:Real, <:Real} = (0.04, 0.15), HF::Tuple{<:Real, <:Real} = (0.15, 0.4))

Calculate the BRS indices αLF, αHF using AR spectral decomposition.

Keywords

  • p: order of the AR process, if a UnitRange is given, the optimal order will be chosen through AIC, defaults to 12:18.
  • nfreq: resolution of the analysed spectra, defaults to 256
  • sf: sampling frequency in Hz, defaults to 1
  • LF: Tuple that defined the LF band, defaults to (0.04, 0.15)
  • HF: Tuple that defined the HF band, defaults to (0.15, 0.4)
  • verbose: print some information oder selection, defaults to false
source
Cardio.BRS.arBRS — Type

Struct that stores all information regarding the arBRS estimation. The final result is stored in αLF & αHF. It can be plotted for visual inspection.

source
result = BRS.arbrs(input.RR, input.SBP; sf = 1000 / mean(input.RR))
plot(result, dpi = 120)

Phase-Rectified Signal Averaging Method

Source: Bauer, A., Morley-Davies, A., Barthel, P., Müller, A., Ulm, K., Malik, M., & Schmidt, G. (2010). Bivariate phase-rectified signal averaging for assessment of spontaneous baroreflex sensitivity: pilot study of the technology. Journal of electrocardiology, 43(6), 649-653. DOI

Cardio.BRS.prsabrs — Function
prsabrs(RR::Vector{<:Real}, SBP::Vector{<:Real}; L::Int = 15)

Calculate a measure for BRS based on phase-rectified signal averaging as defined by Bauer et al. 2010.

Keywords

  • L: defines the segment length as 2L+1
source
Cardio.BRS.prsaBRS — Type

Struct that stores all information regarding the prsaBRS estimation. The final result is stored in value or valueNorm for a normalized value. It can be plotted for visual inspection.

source
result = BRS.prsabrs(input.RR, input.SBP)
plot(result, dpi = 120)