Spatial mapping of respiratory related evoked responses using wavelet and Laplacian transform methods

The authors investigate how mid-latency cortical evoked potentials measured on the scalp are affected in response to pressure pulses of varying magnitude at the mouth. The topology of the radial current produced by these respiratory-related evoked potentials (RREPs) at 30 points on the scalp overlying the somatosensory cortex is quantified using a reference independent technique. The authors computed scalp potential and current density profiles as a function of time post-stimulus after preprocessing the RREPs using wavelet transform methods. Stimulation of peripheral mechanoreceptors has long been known to generate small, short latency potentials in the electroencephalogram (EEG) that reflect the character of the applied stimulus. Previous studies suggested that small magnitude waves can be evoked in the cortical EEG by occluding the inspiratory airway of normal subjects, thereby creating a negative airway pressure that provided a distortional stimulus to respiratory mechanoreceptors. Here the authors apply a novel approach based on wavelet and Laplacian transform methods to the analysis and characterization of respiratory-related evoked potentials. Tile analysis of the scalp potentials was done using the wavelet transform and Laplacian estimates to examine the dynamic change of the RREfs activity with time by sequential analysis of spatial activity. The wavelet shrinkage algorithm was used to provide temporal (1-D) filtering to reduce background noise and EMG noise as much as possible before the estimates of the spatial mapping of the Laplacian of the RREPs. The authors´ results suggest that the activities in the spatial mappings of short (40 trial) averages at three different scales (scale 4: 62.5-125 Hz, scale 5: 31.25-62.5 Hz and scale 6: 15.62-31.25 Hz) accounted for virtually all the information in the RREP response. Estimates of the RREP responses at these scales were significantly increased when the pressure pulse was increased from -5 to -10cm H₂O. Increasing the pressure pulses above -10cm H₂O did not cause any appreciable changes in the spatial mappings at these three frequency scales