Combining EEG and Diffuse Optical Imaging : A New Approach to Monitoring Neonatal Seizures ?

Correspondence to: Dr Topun Austin, Neonatal Intensive Care Unit, The Rosie Hospital, Cambridge University Hospitals, NHS Foundation Trust, Hills Road, Cambridge CB2 2QQ, UK. Email: topun.austin@ addenbrookes.nhs.uk Tel: 01223 216677 Fax: 01223 0217064. Seizures in the newborn infant brain represent a major unsolved problem in neonatal medicine. Neonatal seizures are poorly classified, frequently under-diagnosed, and are difficult to treat. Estimates of incidence vary; a recent review placed the figure at between 1.8 and 3 per 1000 live births, however, the risk for babies with a very low birthweight (< 1.5 Kg) is thought to be higher, between 10-11 per 1000. In the term infant, hypoxic-ischaemic brain injury is the most common cause of seizures, although perinatal stroke is an increasingly recognized cause in the population group. In the preterm infant, haemorrhagic and ischaemic lesions as well as infection account for the majority of seizures. Neonatal seizures have been associated with adverse neurodevelopmental outcome, but there is continuing debate about the extent to which seizures can cause or aggravate brain injury themselves. Traditionally, neonatal seizure has been diagnosed on the basis of clinical observation of changes in behavioural state, motor and autonomic function. It is now known that diagnosis by clinical observation alone dramatically underestimates the prevalence of neonatal seizure. This is because the majority of seizures, as diagnosed by EEG, do not manifest themselves clinically. Video-EEG is rightly considered to be the gold standard for seizure detection, but even with EEG diagnosis can remain difficult because of the high-amplitude, discontinuous nature of neonatal EEG activity. It is also important to consider the case where classic clinical manifestations of seizure are apparent whilst the EEG remains seizure negative. Many of these events can be explained by paroxysmal ‘nonepileptic’ behaviours, but as EEG is known to have limited sensitivity to sub-cortical neurons there remains an important question: do neonates who exhibit clinical manifestations but lack electrographic expression of seizure have a sub-cortical seizure focus, which EEG cannot detect?