Brain Tumour Classification Using Deep Learning

Résumé: A brain tumour is a fatal disease affects children and adults the disease might be detected using physical exam, neurological exam but for the classification, it is done with biopsy. That last one is concerned with brain surgery, which is so hard and complicated itself. Nowadays it is so important for the early detection because of the five-year rate of survival. The early detection and classification could help to choose the perfect plan for treatment. With the big development and change in technology and AI techniques could help in diagnosis and classification without any huge risks, Using the available data of MRI images that are studied from the radiologist. In our study, we took two approaches, the first including four transfer learning models and the second including a CNN model, to both classify different types of brain tumour. Using three different datasets available at kaggle With the CNN approach, we manage to achieve an accuracy of 99.76 %. The Experimental Results shows that our proposed Convolution Neural Network model (CNN) gives the best accuracy as compared to other transfer learning techniques. At last, we make a small comparison with the state of the art methods.

Publié dans la revue: