Maayke Hunfeld

234 Chapter 8 features are associated with outcome after CA (15-17). Extrapolating adult findings to children is inappropriate, since children have age-dependent anatomy and physiology and responses relevant to injury of central nervous system (18). Thus far, only one study has reported the ability QEEG features to predict neurological outcomes in 87 children after CA (19). Their model that best predicted unfavorable neurological outcome had a specificity of 0.75 and positive predictive value (PPV) of 0.79, which illustrates that at this stage it should not be used solely in neuroprognostication. Recently a collaboration has been set up between our PICU, our neurology/ neurophysiology department and Delft University of Technology, with the purpose of integrating machine learning with clinical patient data. One of the next steps is designing a machine-learning algorithm for cEEGs in children of different age groups after OHCA. The primary goal of this project is to investigate the role of QEEG in neuroprognostication after OHCA. Potential solution 1c. Qualitative and quantitative MRI In analogy to potential solution 1a, a scoring system should be designed to report brain MRIs. Preferably, brain MRIs must be performed using to the same acquisition parameters, scan quality (preferably on the same scanner), and timing. The use of quantitative MRI, such as brain diffusion tensor imaging (DTI), might be another prognostic tool for pediatric comatose OHCA survivors. DTI enables to calculate the fractional anisotropy, which indicates the quantification of white matter injuries that occur during and around global anoxia (20, 21). In a prospective multicenter cohort of adult patients who had been comatose for 7 days after CA, whole brain white matter fractional anisotropy could accurately predict neurological outcome at 6 months post-CA (22). Another DTI technique is white matter tractography, which studies the organization of structural connectivity (23). In children with moderate and severe traumatic brain injury (TBI), tractography has revealed abnormal organization of the structural connectome (i.e., the comprehensive map of neural connections in the brain), which was associated with impaired neurocognitive function (24). Up to now, age-dependent normal values of DTI for children are lacking. The Generation R study, a large prospective cohort study in healthy children in Rotterdam from fetal life until young adulthood, might generate these normal values. In this cohort, MRIs are made at the ages of 9 and 13 years (25).