Coherence Between Spontaneous Eeg and Doppler Velocimetry of Cerebral Blood Flow in Newborns During Quiet Sleep

Abstract

Cerebral blood flow (CBF) changes in the newborns (NB) can lead to brain damage by decrease in the oxygen and glucose supply. Aiming at contributing to a better understanding of the underlined mechanisms involved, association between the electroencephalogram (EEG, right front-temporal EEG derivation) and the Doppler velocimetry of the middle cerebral artery from to term NB has been investigated. These signals were simultaneously collected from 20 NB and then epochs during quiet sleep (QS patterns: Tracé Alternant, TA, and High Voltage Slow, HVS) were selected. EEG power parameters within frequency bands of interest, e.g. power in theta (P within 4-8 Hz) and the velocimetry of CBF were estimated at each second. In order to investigate the association in the frequency domain, firstly the cross correlation function (CCF) with missing samples were estimated and then the magnitude square coherence (MSC) calculated. During TA, CCF between P and V resulted in the maximum value of the median close to 0.24 around -5 s (P anticipated to V) for 84.6% of the NB with p < 0.05 (statistical significance was tested with Monte Carlo simulation). The maximum of the MSC median occurred around 0.10 Hz in 92.3% of the NB (p < 0.05). A distinct behavior was observed for HVS. This could mask the results of TA when both patterns are taken together (QS). These findings indicate a significant association between the neuronal activity and CBF during TA. The high coherence value could be interpreted as TA influencing CBF or the existence of another physiologic variable influencing both, the CBF and the neuronal activity.