A Unifying Method to Study Respiratory Sinus Arrhythmia Dynamics Implemented in a New Toolbox

Figures

• 

  Figure 1.. Pipeline and features extraction. A, Overview of the pipeline. The dark green and yellow boxes highlight the uniqueness of the toolbox. B, Respiratory features. All features are individually collected for each cycle. C, RSA features detection. All features are individually collected for each cycle.

• 

  Figure 2.. Computational procedure for RSA extraction. A, Respiratory cycle detection. Baseline crossing points are detected on preprocessed signal: inhalation and exhalation starts and represented by green and red dots, respectively. B, R peak detection on ECG. After ECG filtering, R peaks are detected by finding local maxima (purple dot). C, Instantaneous Heart rate signal reconstruction. RR intervals in seconds are converted to beats per minute. RR intervals are interpolated on regularly sampled time series. Respiratory epochs are displayed by pink and green time zones for exhalation and inhalation respectively. D, Cyclical deformation of heart rate epochs. Heart rate signals are windowed according to respiratory cycle epochs. Each epoch time axis is rescaled to a respiratory phase basis with alternating inhalation and exhalation phases. E, Respiratory cycles stretched. 0: inhalation starting phase point; 0.4: inhalation-exhalation transition phase point; 1: exhalation stop phase point. The average waveform is plotted in orange. F, RSA dynamics along respiratory phase. With the same process, heart rate dynamics are computed and plotted along respiratory phase. Various heart rate epochs…

• 

  Figure 3.. Some examples of respiratory phased RSA use. A, Intersubject variability of the mean RSA. Subjects are color coded. Shades correspond to 1 SD. B, Intrasubject variability of the RSA. Cycle-by-cycle RSA of one subject are represented. Color palette, from dark to light warm colors, encodes the amplitude of the corresponding inhalation (from dark to yellow). The subject corresponds to the one colored in dark gray in Figure 1A. C, RSA to inhalation volume correlations. Only significant slopes are represented.

• 

  Figure 4.. Examples of processing rat dataset. A, Respiratory signal. Fast breathing periods are displayed by a shadow color zone. B, ECG signal. ECG peaks have been detected thanks to rodent preset of parameters. C, Heart rate. D, Respiratory phased RSA average. Average of all RSA cycles while splitting the respiratory cycle in “fast” (<600 ms) and “slow” (>600 ms). Shadow represents 1 SD on each side.