Cardiovascular disease (CVD) has emerged as the primary cause of global mortality in recent years, often manifesting initially with mild symptoms that escalate gradually. Diagnostic tools such as cardiac magnetic resonance imaging (CMRI) are pivotal in identifying CVD. However, interpreting CMRI scans can be challenging due to their extensive and inconsistent data. This research addresses these challenges by investigating the application of deep learning, specifically leveraging the U-Net architecture, for automated segmentation of CMRI images. The study aims to classify patients into categories such as normal, hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), and arrhythmogenic categories like right ventricle (ARV), among others. Using the MICCAI dataset, the study demonstrates the efficacy of convolutional neural networks (CNNs) in CVD diagnosis. Notably, the study achieves Dice Index scores of 0.873, 0.837, and 0.846, and Jaccard Coefficients of 0.857, 0.819, and 0.814 in the test set for LV, RV, and MYO regions, respectively. Furthermore, features extracted using Pyradiomics are utilized for classification using classifiers such as Random Forest (RF) and Gradient Boosting Machine (GBM).