Postanoxic coma is caused by global anoxia of the brain, most often due to cardiac arrest. Brain monitoring with electroencephalography (EEG) has been shown to provide prognostic information for recovery. Previous studies showed changes in connectivity between cortico-cortical and cortico-subcortical regions in pathological and pharmacological coma patients. While synchronous EEG activity is linked to a high cortico-cortical connectivity, arousals can be measured in different EEG channels and are linked to high cortico-subcortical connectivity. Both reflect brain dynamics and provide informations for recovery from postanoxic coma. As part of the PhysioNet Challenge 2023, we (ibmtPeakyFinders) propose a novel approach to estimate the degree of anoxic brain injury from cardiac arrest and the likelihood of recovery by evaluating brain dynamics. We used the time delay stability method to assess the coupling between different EEG channels reflecting cortico-cortical connectivity. Furthermore, we applied a self-developed arousal detector for the assessment of cortico-subcortical connectivity. Additionally, we extracted features from all EEG channels via a convolutional neural network to characterize brain activity. Features of the three different types are calculated from 5-minute segments at different time steps within the first 72 hours after patient admission. To monitor the development of brain activity over time, the features of different time steps are combined into a multivariate time series and processed by a long short-term memory network to generate sequence features. These features are used together with the patient metadata to predict the outcome of postanoxic coma patients. A first test implementation of our approach achieved a true positive rate of 0.42 at a false positive rate of 0.05 on the hidden test dataset. We hypothesize that patients with higher brain dynamics have a higher chance of recovery from post-anoxic coma. By evaluating feature importance, we expect new insights in the interaction of brain dynamics from EEG.