Across all stages of life, infants are amongst the most susceptible to brain injury. This vulnerability is particularly pronounced in a condition called Hypoxic Ischemic Encephalopathy (HIE), where a lack of oxygen during the birthing process leads to encephalopathy and neuronal cell death. Clinicians rely on neuroimaging modalities, particularly electroencephalography and magnetic resonance, during the first week of life to assess HIE severity and detect and manage seizures that result from HIE. This body of work focuses on extracting features from neuroimaging modalities to prognosticate infant outcomes and detect seizures. In the first part of the project that focuses on prognostication, it was found that a combination of cortical insula injury and absence of EEG state change activity could predict abnormal outcomes in neonates between three and six months of life. It was also found that periventricular white matter injury correlated significantly with abnormal seizure activity and poor outcomes in term neonates. In the second part of the project that focuses on seizure detection, it was found that a combination of features from the amplitude-integrated EEG and compressed spectral array could detect seizures with high accuracy using an external dataset of 79 patients with heterogenous disease etiology. Results also revealed that the algorithm performed more poorly on HIE patients within the dataset. A follow-up algorithm validation study using the inhouse cohort of term HIE infants showed that the algorithm could perform well on an independent HIE cohort and equally as well if only trained using the subset of external data patients that had HIE due to birth asphyxia. In conclusion, this body of work shows that early clinical management for neonatal patients can be improved using a combination of features from EEG and MRI to score outcomes and a combination of quantitative EEG features to detect seizures better.