Classifying brain images and diagnosing tumors using MRI is a difficult process due to the diversity and complexity of the cases. This research presents a model for classifying brain images and detecting tumors based on deep learning techniques.
The proposed model has been trained and tested on a database of MRI images consisting of 253 images. 155 of the images are images of healthy cases and 98 image are images of defected cases. The images have been pre-processed and prepared as suitable input for the designed neural network model. The proposed model has achieved outstanding results where the accuracy reached 94.5% with an AUC value = 0.95. The obtained results make the proposed model a suitable model to help specialist doctors in diagnosing brain images and identifying cases of malignant tumors.