8 A Self-Organized Rhythm in Peripheral Effectors: The Intermediary …
141
8.2.1 Data Acquisition
All physiological data were recorded on-line with a personal computer equipped with
an analog-to-digital (AD) converter (Data Translation, DT 2812 a), as well as a software package (DiaDem®, National Instruments). Data were recorded at a sampling
rate of 300 Hz with 12-bit resolution. They were stored for further processing on
digital media.
We used a photoplethysmographic probe (PPG) which was attached to the glabella
forehead region known to display little or no vasoconstrictor activity at room temperature for this could have altered rhythms in these delicate signals. Irrespective of
certain interpretative deficiencies regarding the amplitude of these signals, sufficient
precision in the temporal domain of these signals was demonstrated by simultaneous recording of PPG and LASER-Doppler [21]. To detect respiration-induced
fluctuations contained in cardiovascular signals, a belt equipped with a strain-gauge
(ADMS) was used for the registration of the frequency and amplitude of respiratory
induced thoracic movements.
8.2.2 Data Analysis
Data were analyzed off-line using software designed for the non-linear analysis of
multiple time series of bio-physiological data (commercially available via Procalysis®, Simplana GmbH/Aachen-Germany). Slow trends in DC-offset were removed.
The following nonlinear algorithms were applied to data sets.
Multiscaled Time-Frequency Distribution (mTFD)
The changes of frequencies contained in physiological signals can be ideally analyzed
using Morlet-wavelet based multiscaled Time–Frequency-Distribution (mTFD) as
this method combines analysis of temporal information concerning fluctuations in
both frequencies and amplitudes [18, 22].
For mTFD computations, the interval length analyzed comprised usually the entire
recording time with an average of approx. 180.000 samples. The frequency range
was 0.04–0.6 Hz which equaled period length from 25 to 1.66 s. As each window
contained 7 periods, window lengths were from 11.66 to 175 s. With frequency and
time resolution set in general to 200, this resulted for an epoch of 600 s an overlap
for the lowest frequency of 172 s, for the highest frequency of 8.66 s. This overlap
helped smooth spectra within the mTFD.
141
8.2.1 Data Acquisition
All physiological data were recorded on-line with a personal computer equipped with
an analog-to-digital (AD) converter (Data Translation, DT 2812 a), as well as a software package (DiaDem®, National Instruments). Data were recorded at a sampling
rate of 300 Hz with 12-bit resolution. They were stored for further processing on
digital media.
We used a photoplethysmographic probe (PPG) which was attached to the glabella
forehead region known to display little or no vasoconstrictor activity at room temperature for this could have altered rhythms in these delicate signals. Irrespective of
certain interpretative deficiencies regarding the amplitude of these signals, sufficient
precision in the temporal domain of these signals was demonstrated by simultaneous recording of PPG and LASER-Doppler [21]. To detect respiration-induced
fluctuations contained in cardiovascular signals, a belt equipped with a strain-gauge
(ADMS) was used for the registration of the frequency and amplitude of respiratory
induced thoracic movements.
8.2.2 Data Analysis
Data were analyzed off-line using software designed for the non-linear analysis of
multiple time series of bio-physiological data (commercially available via Procalysis®, Simplana GmbH/Aachen-Germany). Slow trends in DC-offset were removed.
The following nonlinear algorithms were applied to data sets.
Multiscaled Time-Frequency Distribution (mTFD)
The changes of frequencies contained in physiological signals can be ideally analyzed
using Morlet-wavelet based multiscaled Time–Frequency-Distribution (mTFD) as
this method combines analysis of temporal information concerning fluctuations in
both frequencies and amplitudes [18, 22].
For mTFD computations, the interval length analyzed comprised usually the entire
recording time with an average of approx. 180.000 samples. The frequency range
was 0.04–0.6 Hz which equaled period length from 25 to 1.66 s. As each window
contained 7 periods, window lengths were from 11.66 to 175 s. With frequency and
time resolution set in general to 200, this resulted for an epoch of 600 s an overlap
for the lowest frequency of 172 s, for the highest frequency of 8.66 s. This overlap
helped smooth spectra within the mTFD.