Brain tumors identification and classification is a most crucial role in medical diagnosis and treatment plan for the patients. The information acquired by medical imaging machines contains low level information which cannot be perceived my human visual system. An automatic Computer aided diagnosis system with deep convolutional networks can perform well in Medical Image classification, but requires large datasets. Generally it is difficult to obtain a large number of labelled samples in medical image classification tasks. Knowledge transfer with fine tuning of pre-trained networks can apply on small medical datasets for diagnosis. In this work transfer learning and block wise fine tuning with separate KNN classifier on pre trained deep neural network VGG net is used to classify BRATS and CE-MRI datasets. The proposed method is denoted as BT-VGGNet. The labelled BRATS Flair Images and CE-MRI T1-weighted contrast-enhanced datasets are exploited to block wise fine-tune all hidden layers in the VGG deep neural network in feature extraction and KNN for classification. Experimental results show that the proposed model exceeds the state-of -the-art classification with 97.28 percent and 98.69 percent accuracy on the BTDS-2 and CE-MRI datasets respectively.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.