Implementation of a Deep Neural Network for Classifying Images of Brain Tumors

Purpose: Identification and categorization of brain tumors is a cyclical process in which tumor components are assessed and suggestions for therapy are made based on their classifications. Many imaging techniques are used for this work. Because MRI provides better soft tissue than CT, and MRI does not involve radiation. The currently available manual method is inefficient and hence we provide an advanced method by using the deep learning concepts. Materials and Methods: This MRI creates detailed images of our body's organs and tissues by using a computer's radio wave and an attracting field. Deep Learning (DL), a subset of Machine Learning (ML), is helpful for the categorization and identification of issues. This project uses one dataset consisting of three categories (Meningioma, Glioma, Pituitary. Results: In this work, the first stage is pre-processing concerning two datasets. Later involves detection by using a Convolution Neural Network algorithm (CNN). The suggested CNN performs admirably, with the greatest overall accuracy for the datasets coming in at 94.3% and 96.1%. The final results demonstrate the model's capability for brain tumor classification and detection problems. Conclusion: The proposed system helps to automatically differentiate between the types of Tumors from the normal, future it can be improved to analyze the brain tumor and classification, which will be more useful in the treatment. A few more sectors of artificial intelligence can also be incorporated along with the proposed system to increase the standard of the proposed system