There have been many prolonged neural and inflammatory diseases which have advanced to be pervasive and life-threatening. These diseases can easily be overlooked, as they often tend to yield minor symptoms, like headache, fatigue or sensory changes. Therefore, early detection of cerebral maladies can be achieved by examining, analyzing, and integrating modern concepts like magnetic resonance imaging (MRI) and machine learning. The classification of abnormalities like Glioma, meningioma, pituitary malfunctioning, or normal conditions can be achieved using a convolutional neural network or CNN. A comparative study of different CNN architectures allows the user to choose the best and most robust design for their application. There can be many varying parameters, which can fluctuate when compared to the other readily available ones.