High precision brain tumor classification model based on deep transfer learning and stacking concepts

In this article, we proposed an intelligent clinical decision support system for the detection and classification of brain tumor from risk malignancy index (RMI) images. To overcome lack labeled training data needed to train convolutional neural networks, have used a deep transfer learning stacking concepts. For this, choosed seven networks (CNN) architectures already pre-trained on ImageNet dataset that precisely fit magnetic resonance imaging (MRI) tumors collected segmentation (BraTS) 19 database. improve accuracy our global model, only predict as output prediction obtained maximum score among predictions CNNs. We 10-way cross-validation approach assess performance main 2-class model: low-grade glioma (LGG) high-grade (HGG) tumors. A comparison results model with those published in literature, shows is more efficient than average test precision 98.67%, f1 98.62%, 98.06% sensitivity 98.33%.