Multicentre Clinical evaluation of a neonatal seizure detection algorithm. (360G-Wellcome-098983_Z_12_Z)

Multicentre clinical evaluation of a neonatal seizure detection algorithmSeizures are difficult to detect in new-born babies and can be a marker of brain injury, resulting in lifelong disability. Treating seizures early is more successful than treating them at a later point, and there is evidence to show that controlling seizures can reduce brain damage. Seizures cannot be treated unless clinicians have an accurate way of measuring them – recognising seizures by observing the baby alone is not reliable.The only way to accurately detect these events is to monitor electrical brain patterns using wires attached to the scalp (EEG). However, EEG signals are very difficult to interpret and full-time monitoring of EEG outputs by trained clinicians is not practical. Using funding from a Translation Award, Liam Marnane and colleagues at the University College Cork have developed an automated intelligent EEG computer program that can detect seizure accurately. They will now test the system in neo-natal intensive units around the world to show its benefits for new-borns. The goal is to make the software reliable and compliant with international regulatory requirements so that it is suitable for inclusion in multiple medical monitors.