Simultaneous MEG and Ultrasound Recording of Fetal Breathing Movement

Abstract

Event Abstract Back to Event Simultaneous MEG and Ultrasound Recording of Fetal Breathing Movement Umit Ulusar1*, Rathinaswamy Govindan2, James Wilson1, Murphy Pamela2, Curtis Lowery2 and Hari Eswaran2 1 University of Arkansas at Little Rock, Graduate Institute of Technology, United States 2 University of Arkansas for Medical Sciences, Department of Obstetrics and Gynaecology, United States Bio-magnetic signals induced by fetal breathing movement (FBM) are interfering factors for the investigation of fetal magnetoencephalography (fMEG) and magnetocardiography. We developed a novel framework to identify FBM signal in the bio-magnetic recordings. Using a 151 magnetic sensor array system and a portable ultrasound device, simultaneous fMEG and sonographic recordings were performed for 11 to 30 minutes duration on six fetuses ranging between 35 and 37 weeks gestational age. The fMEG data were sampled at a rate of 312.5 Hz using a bandwidth set for 0-100 Hz. The interfering maternal cardiac signals were removed by the signal space projection technique. Ultrasound video was saved in MPEG format with 30 frames per second and 640 x 480 pixel resolution. The ultrasound images were processed using Sticks filter [1] for the elimination of the speckle noise and with horizontal Sobel filter for the edge enhancement. We obtained fetal movement signal by tracing the movement of the boundary defined by the fetal thorax spanning multiple frames. This signal was filtered using a 4th order Butterworth band-pass filter between 0.5 to 10 Hz with zero phase distortion. This resulted in a time series sampled at the video frame rate that represents the temporal behavior of the fetal thorax. Finally, to identify the MEG sensors containing the FBM, we used spectral coherence. The coherence was calculated between the MEG data and FBM signal obtained with ultrasound image processing on a window length of 1 minute with 4-second slides. The fMEG sensors that displayed statistically significant coherence were identified as the ones that contain the FBM signal.This study showed the feasibility of simultaneously recording fMEG and ultrasound. In future, the data observed with the identified sensors can be used to characterize the bio-magnetic signal of the FBM.