Classification of Neonatal EEG
A classification of neonatal EEG is a multiple valued description of the EEG that is capable of defining particular EEG signal characteristics. A typical classification of EEG involves observing specific EEG features such as the continuity, amplitude, frequency content, symmetry and synchrony combined with monitoring particular EEG patterns such as sleep states, spikes and sharp waves. Classification of neonatal EEG is an important first step in developing the application of EEG for diagnosis and prognosis.
We aim to classify the EEG of normal infants and infants suffering from various conditions to see if there are differences in the EEG between these groups than can assist in diagnosis and prognosis. We aim to classify the EEG of;
- normal healthy newborns,
- newborns suffering HIE,
- newborns suffering from seizure.
We have begun by classifying the EEG of normal term infants. These babies show no sign of injury at birth. The reason we are studying normal infants is because very little is known about the EEG patterns of the healthy term newborn baby early after birth as most studies focus on the EEG of sick infants. It is also, impossible to interpret “abnormal” EEG without knowing what normal neonatal EEG looks like.
We have studied 30 healthy full term newborn babies within the first 12 hours after birth. We used early recordings of EEG as the early diagnosis of HIE is crucial when treating with cooling. Video-EEG (F4, T4, P4, F3, T3, P3, Cz), ECG and respiration recordings were commenced as soon as possible after birth and continued for one to two hours to include active and quiet sleep phases. Examples of active and quiet sleep periods are demonstrated in Figure 1.
Figure 1: Active and quiet sleep on the EEG of a healthy full term newborn.
Visual and spectral analyses of all recordings were performed to classify the EEG of the healthy baby very early after birth. Our study has revealed;
- Evidence of continuous symmetric and synchronous activity in all recordings with clearly developed sleep-wake cycles.
- The proportions of active sleep (AS) and quiet sleep (QS) within 12 hours of birth. This is demonstrated in Figure 2.
- Clear differences between spectral measures under active and quiet sleep conditions. This is shown in Figure 3.
Figure 2: Proportions of sleep states in one hour of recording within 12 hours of birth.
Improvements in the classification of EEG have the potential to improve the diagnostic and prognostic ability of the EEG.
Figure 3: Differences in spectral EEG measures during active and quiet sleep.
Further information regarding the classification of EEG in the newborn can be found in,
- I. Korotchikova, S. Connolly, C.A. Ryan, D.M. Murray, A. Temko, B.R. Greene and G.B. Boylan. "EEG in the healthy term newborn within 12 hours of birth". Clinical Neurophysiology, doi:10.1016/j.clinph.2009.03.015
- R. Pressler, G. Boylan, M. Morton, C. Binnie, J. Rennie. "Early serial EEG in hypoxic ischaemic encephalopathy". Clinical Neurophysiology, 112, pp. 31-37, 2001.
- G.B. Boylan, J.M. Rennie, and D.M. Murray. "The normal neonatal EEG" in Neonatal Cerebral Investigation, Chapter 6, Eds: J.M. Rennie, N.J. Robertson and C.F. Hagmann. Cambridge University Press, UK, 2008.