Brain Tumor Classification In MRI Images Using Convolutional Neural Network

Brain tumor is a severe cancer disease caused by uncontrollable and abnormal partitioning of cells.It has recently been a leading cause of death for many people.Of all tumors,the severity of a brain tumor is very high.Recent progress in the field of deep learning has helped the health industry in Medical Imaging for Medical Diagnostic of many diseases.To examine a tumor disease in the brain,lungs,prostate,liver or breast different image technologies such as computed tomography(CT),ultrasound images and magnetic resonance imaging(MRI),are used.For Visual learning and Image Recognition,task CNN is the most prevalent and commonly used machine learning algorithm.Similarly,in our paper,we introduce the convolutional neural network(CNN)approach along with Data Augmentation and Image Processing to categorize brain MRI scan images into cancerous and non-cancerous.First we applied image processing technique by cropping out the region of interest then in order to prevent overfitting we applied data augmentation on the preprocessed MRI images to increase the size of our dataset.Then we trained it through our simple proposed CNN model having only 8 hidden layers and by using Adam optimizer to reduce the loss in our model.Similarly using the transfer learning methodology with the same approach having the same MRI scan dataset we compared the accuracy performance of our scratched CNN model with pre-trained VGG-16,ResNet-50,and Inception-v3 models.As the experiment is tested on a very small dataset but the experimental result shows that our model accuracy result is very effective and have very low complexity rate by achieving 100% accuracy,while VGG-16 achieved 96%,ResNet-50 achieved 89% and Inception-V3 achieved 75% accuracy.Moreover Precision,Recall,F1 score and ROC curve of proposed CNN model was also evaluated and compared with other pre-trained models.So the experimental result shows that our model requires very less computational power and has much better accuracy results as compared to other pre-trained models.