Knowledge of whole heart anatomy is a prerequisite for many clinical applications. For obtaining precise measurements of clinical metrics for the substructures of the heart, their accurate modeling in a 3D setup is requisite. In this paper, we adopt a clear-cut segmentation architecture with adequately proven results for the automatic segmentation and assessment of the heart and its substructures from contrast-enhanced cardiac CT (computed tomography) scans. A two-step framework consisting of a U-net and an end-to-end trained Spatial-Configuration net is considered for the preliminary segmentation task. This network annotates seven different mediastinal structures: the left ventricle, right ventricle, left atrium, right atrium, ascending aorta, pulmonary artery, and myocardium. By achieving the average Dice score of 86.73% concerning manual annotations made by a trained expert, we focus on medical assessment by developing a tool to measure the essential cardiac functionalities: RA to LA ratio, RV to LV ratio, and the radii of Cardiac Vessels. MRAE for RV:LV and RA:LA calculations over the test set are 10.10% and 10.40%, respectively. For radial measurements of the aorta and pulmonary artery, the obtained MRAE are 5.77% and 12.39%, respectively. © 2022 IEEE.