This chapter explores the intersection of deep learning and Magnetic Resonance Imaging (MRI)-based analysis to advance our understanding of brain disorders. By discussing cutting-edge deep learning algorithms, we introduce various approaches alongside MRI modalities to improve the detection, classification, and interpretation of adult brain disorders and neonatal hypoxic-ischemic encephalopathy (HIE). Key methodological approaches include MRI classification, segmentation, reconstruction, and registration, each contributing uniquely to the analysis of both adult and infant brains. For HIE, we discuss challenges posed by limited data, complex lesion characteristics, and the potential of machine learning to enhance prognosis and neurodevelopmental outcome prediction. Additionally, we introduce brain age prediction as a biomarker for abnormal brain aging, highlighting its role as a predictive tool for various brain disorders.