Brain Tumour MRI Image Segmentation and Classification Using FCNN

Abstract

AbstractBrain tumours are one of the most prevalent and aggressive diseases across the globe, with a relatively short life expectancy in their most severe form. Thus, identification and treatment of brain tumour play an important role in improving the patients’ quality of life. In general, the biomedical imaging methods such as computed tomography (CT), magnetic resonance imaging (MRI) and ultrasound images are used to assess tumours in different parts of human body such as brain, breast, lungs, liver, etc. Amongst the diagnostic technologies, magnetic resonance imaging is the most popular imaging approach offered by experts to analyze brain tumours. At the same time, image segmentation also plays a fundamental role in brain tumour analysis and classification. Recently, various image segmentation approaches are used to characterize the brain tumour in the early stages of development. However, the massive amount of data collected by an MRI scan makes manual categorization of tumour and non-tumour in a given period is impossible, and it also has several limitations (for example, reliable quantitative values are supplied for a restricted number of images). As a result, a reliable and automated image segmentation and classification technique are required to reduce human mortality. Automatic brain tumour detection using fully convolutional neural networks (FCNNs) classification is suggested in this paper. When compared to all other state-of-the-art technologies, experimental findings demonstrate that the proposed FCNN achieves a good accuracy with reduced complexity.KeywordsIntracranialConvolutionBrain tumourFCNNSegmentationClassificationAbstraction