Medical imaging is progressing more quickly with the help of cutting-edge research and developments. Artificial Intelligence, deep learning, and virtual reality are key technologies of modern imaging applications. Hence, it is important to identify the necessary application level gaps and to develop a strategic model for performance improvements. One of the most prominent areas of medical imaging is brain tumor identification at an early stage. This paper presents the pre-and post-operative brain tumor segmentation for standard MR image scans. Also, it facilitates the newly developed "CdCnet" algorithm logic for contrast dye injection MR image scans. The brain tumor borders' visibility is less during and after the brain surgery due to tiny veins, which may impact the human body's functioning if it gets damaged during the surgery due to less visibility. Hence, contrast dye injection into the bloodstream helps the system to see arteries and veins. The proposed research methodology can be a new development for segmenting brain tumors of any grade