Deep Learning-Based Automated Detection and Classification of Brain Tumor with VGG16-SVM in Internet of Healthcare

Emergence of deep neural networks in the healthcare has transformed the process of analyzing medical images especially when it comes to diagnose brain tumor disease. As patients are asked to undergo computed tomography, magnetic resonance imaging, etc. following traditional approaches by medical experts which consume a lot of time and abnormal growth of tissue inside brain may remain undiagnosed due to its extremely tiny size at initial stage. Therefore, developing an automated system utilizing artificial intelligence, deep learning in internet of healthcare can overcome drawbacks associated with traditional approaches to achieve efficient, accurate and quick outcomes in the healthcare to save billions of lives worldwide. Consideration of 3264 brain MRI images has been done which have been acquired from kaggle comprising of 2764 images having tumor and 500 healthy brain MRI images. In this paper, a novel approach has been presented comprising of a pre-trained model namely visual geometry group having 16 layers, a well-established convolutional neural network to extract significant features from input data which are further fed to the support vector classifier for distinguishing infected images from health ones. A process of transfer learning has also been deployed with VGG16 due to acquiring its pre-trained features from keras to save a lot of time while training the proposed model. Moreover, Internet of Healthcare framework can aid radiologists in making decisions on real-time applications to provide timely recommendation as well as treatment to the patients. Thus, validation is performed on unseen data to ensure efficient performance of the model and model gained 98.16% accuracy, 99.09% precision, 98.73% recall and 98.91% F1-score which outperforms the existing approaches and demonstrated potential of revolutionizing neuroimaging field and patient care. © 2023, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.