AIM:To develop and evaluate an algorithm for the automatic screening of electrographic neonatal seizures (ENS) in amplitude-integrated electroencephalography (aEEG) signals. METHODS:CFM recordings were recorded in asphyxiated (near)term newborns. ENS of at least 60 sec were detected based on their characteristic pattern in the aEEG signal, an increase of its lower boundary. The algorithm was trained using five CFM recordings (training set) annotated by a neurophysiologist, observer1. The evaluation of the algorithm was based on eight different CFM recordings annotated by observer1 (test set observer 1) and an independent neurophysiologist, observer2 (test set observer 2). RESULTS:The interobserver agreement between observer1 and 2 in interpreting ENS from the CFM recordings was high (G coefficient: 0.82). After dividing the eight CFM recordings into 1-min segments and classification in ENS or non-ENS, the intraclass correlation coefficient showed high correlations of the algorithm with both test sets (respectively, 0.95 and 0.85 with observer1 and 2). The algorithm showed in five recordings a sensitivity > or = 90% and approximately 1 false positive ENS per hour. However, the algorithm showed in three recordings much lower sensitivities: one recording showed ENSs of extremely high amplitude that were incorrectly classified by the algorithm as artefacts and two recordings suffered from low interobserver agreement. CONCLUSION:This study shows the feasibility of automatic ENS screening based on aEEG signals and may facilitate in the bed-side interpretation of aEEG signals in clinical practice.

The cerebral function monitor (CFM) is a device to monitor the cerebral function of newborns at the neonatal intensive care unit (NICU). This relatively new device is used by nurses, residents and doctors of the neonatology department. Their experience in these types of signals is limited and training is needed for an optimal use of the monitor in performance and interpretation of the signals. In the training program design of this screen based simulator, we address subsequently a theoretical introduction, placement of the electrodes, the quality of the signal, recognition of artefacts and interpretation of signals of specific pathologies. In each section the trainee will be tested by his/her actions and by multiple choice questions. Positive feedback and explanations of answers will be given by the program. The scenarios included in the program are birth asphyxia and movements that could be seizures. In addition, attention will be paid to differences in clinical course of disease and effects of medication, with respect to the occurrence of seizures and change in back ground pattern. Different types of artefacts are included as well. The program features different levels of skills for doctors and nurses as well as a review mode where the theoretical introduction, placement of electrodes and the questions will be left out. Using this program, we aim to improve the adequate use of the CFM and hence increase the quality of care in the NICU.

Background: From clinical point of view it is important to detect all seizure activity in neonates. The cerebral function monitor (CFM) processes the electro encephalogram (EEG) into an amplitude-integrated EEG (aEEG) signal that is relatively easy to read and interpret compared to the classical EEG. So far seizures were defined based on the visual interpretation of a typical pattern. Objective: The objective of this study is to create a program for automatic detection of seizures in neonatal aEEG signals, based on a quantitative analysis. Seizures of a length of more than 60 seconds are included. Method: Since high cerebral activity corresponds with high amplitude in the aEEG signal, the program detects rises in amplitude of the lower margin, compared to the background signal. The lower margin is defined as a threshold value, where 95% of the aEEG samples are above this threshold. If this lower margin for a segment is significantly higher than the mean lower margin of the 6 minutes prior to the segment, and if this difference in height continues for at least 60 seconds, this pattern is marked as a seizure. Muscle artefacts are detected based on the frequency content of the EEG signal. Other artefacts are detected using the amplitudes of aEEG, EEG and impedance of the electrodes. Three signals were used to train the program. Results: Three CFM recordings of 9 hours from full-term newborns with different background patterns were used to evaluate the program. The signals were annotated by an expert in neonatal neurophysiology, who found in total 95 seizures. The program detected 101 patterns as seizures, with a sensitivity of 98% (93 true positives) and a positive predictive value of 92% (8 false positives). Conclusion: This study indicates that it is possible to automatically detect seizures of more than 60 seconds, using CFM recordings.