Quantifying Fetal Evoked Responses using Spectral Coherence Measure

Abstract

Event Abstract Back to Event Quantifying Fetal Evoked Responses using Spectral Coherence Measure Rathinaswamy Govindan1*, Hubert Preissl2, 3, Pamela Murphy1, Maureen Ware1, Curtis Lowery1 and Hari Eswaran1 1 University of Arkansas for Medical Sciences, Department of Obstetrics and Gynaecology, United States 2 University of Tubingen, MEG-Center,University Hospital, Germany 3 University of Arkansas, Department for Obstetrics and Gynecology, Medical College, United States Traditionally, brain evoked responses (ER) are quantified by performing event-related averaging. However, this approach has limitations if applied to fetal brain signals to quantify the ER because of their small amplitude and interference from other signals such as fetal movements. In order to quantify the fetal ER, one of the following approaches can be taken: (i) non-parametric approach such as bootstrap to assess the significance of ER (ii) use of parametric approach. Here we propose to use spectral coherence to quantify ER. For this purpose we collect auditory ER (AER) using a 151 SQUID array system. Using signal space projection technique, we attenuate the interfering maternal and fetal cardiac signals. To this end, we compute the spectral coherence (SC) between the trigger series and the fetal magentoencephalogram (fMEG) signals. SC is computed as the ratio of the estimate of the cross-spectrum between the trigger signal and fMEG to product of the auto-spectra of the two signals. The value of coherence above the pre-defined significance level is considered as significant coherence (i.e. AER). For the sensors displaying significant coherence the latency of the ER is computed using the phase spectrum, which is defined by the cross-spectrum. Latency is computed by normalizing the phase value in the frequency band of significant coherence by that frequency itself. In this work we demonstrate the application of this approach to serially recorded AER from fetuses at different gestation age and discuss their neurological maturation.