A Machine Learning Approach for Brain Tumor Detection Using CNN Algorithm

Designing automated systems for extracting and identifying brain tumors, including their specific types, from Magnetic Resonance Images (MRIs) is one of the most researched areas of this era. Brain tumors are a crucial cause of human death every year, with around 50% of diagnosed patients succumbing to primary brain tumors worldwide annually. Healthcare professionals use electronic modalities to diagnose brain tumors, and Magnetic Resonance Imaging is one of the most widely used and popular modalities for brain tumor diagnosis. While 2D MRI provides a restricted view of the brain, 3D MRI offers a comprehensive perspective. This work aims to create an automatic system for swift brain tumor diagnosis, especially in critical situations. This system utilizes MRI tests and image processing techniques such as segmentation to identify the type of tumor. The image captured after the slide test undergoes processing to detect the occurrence and specific types of brain tumors. Subsequently, the classification algorithm, including Machine Learning algorithms, determines the tumor analysis. Finally, all the relevant information will stored in a database. The system demonstrates high accuracy in tumor segmentation and type identification, as evidenced by the results.