Transfer Learning Using Convolutional Neural Network Architectures for Brain Tumor Classification from MRI Images

Part 3: Image processing International audience Brain tumor classification is very important in medical applications todevelop an effective treatment. In this paper, we use brain contrast-enhanced magneticresonance images (CE-MRI) benchmark dataset to classify three types of brain tumor(glioma, meningioma and pituitary). Due to the small number of training dataset, ourclassification systems evaluate deep transfer learning for feature extraction using ninedeep pre-trained convolutional Neural Networks (CNNs) architectures. The objectiveof this study is to increase the classification accuracy, speed the training time andavoid the overfitting. In this work, we trained our architectures involved minimal pre-processing for three different epoch number in order to study its impact onclassification performance and consuming time. In addition, the paper benefitsacceptable results with small number of epoch in limited time. Our interpretationsconfirm that transfer learning provides reliable results in the case of small dataset. Theproposed system outperforms the state-of-the-art methods and achieve 98.71%classification accuracy