Electrocortical activity in the near-term ovine fetus: automated analysis using amplitude frequency components

Abstract

We have designed an automated method for analyzing electrocortical (ECoG) activity in the near-term ovine fetus to process and quantitatively classify large amounts of data rapidly and objectively. Seven chronically catheterized fetal sheep were studied for 8h each at ~0.9 of gestation with continuous recording of ECoG activity using a computerized data acquisition system. Multiple ECoG amplitude and frequency parameters were scored from which we established animal specific parameter cut-off values as well as population based duration cut-off values to distinguish low-voltage/high frequency (LV/HF) and high-voltage/low frequency (HV/LF) state epochs, and indeterminate voltage/frequency (IV/F) and transition period activities. We have shown that the incidence of the predominant LV/HF and HV/LF activity states at 45% and 36% of the time, respectively, is comparable to that previously reported using semi-quantitative techniques with visual analysis. However, the duration of these state epochs is considerably shorter due to the detection of brief periods of IV/F activity which would be difficult to capture using visual analysis. Importantly, our findings in the healthy ovine fetus near-term using this automated ECoG scoring methodology now provide a framework from which to study maturational events in younger animals, and under adverse pregnancy conditions.