Vascular Diseases are the leading cause of death. One can mitigate them with healthy habits and medical assistance. Invasive and non-invasive coronary angiography are medical exams that assess the structural and functional status of the vascular system to support physicians with their diagnostics. Therefore, a more customized approach is still needed to reduce inter-observer and intra-observer variability and improve diagnostic consistency. We propose in this paper a process to quantify the characteristics of the human vascular system, considering a hybrid integrated approach. The proposal has five independent modules: 1 - automatic segmentation; 2 - creation of three-dimensional models of the anatomical structures; 3 - modeling the vascular system as a biomedical network; 4 - objective metrics extraction based on network science; 5 - and qualitative inference. We validate our proposal on images of the heart and retinographies. We present that we can extract relevant topological and semantic features in those cases.