Brain Tumor Segmentation and Retrieval Using Deep Learning Techniques

Abstract

Magnetic resonance image (MRI) brain tumor segmentation is crucial and important in the medical field, which can help in diagnosis and prognosis, overall growth predictions, Tumor density measures, and care plans needed for patients. Brain tumor detection and segmentation is an Information retrieval system (IRS) that retrieves information from images. Information retrieval models possess kind of attributes and choosing the best model for application is undoubtedly important. The first aim of this research work is to analyze information retrieval models on an attribute basis. newlineBrain tumor detection and segmentation through images has been taken the application for this research work. This thesis next presented a comprehensive review of state-of-the-art deep learning algorithms applied specifically for automatic and accurate brain tumor segmentation and subsequent evaluation. The datasets included in this study are standard datasets that are widely popular, and known to be part of the Multimodal Brain Tumor Segmentation (BraTS) Challenge that is held annually for the medical domain. This work summarized the performance of deep learning algorithms on BraTS datasets. Algorithms have been compared and summarized against the baseline models with specific attributes and show the performance results for tumor segmentation. The observation is that RescueNet performed best on BraTS 2015 and BraTS 2017 and achieved a high segmentation accuracy of 0.94 dice score, and a high sensitivity of 0.88 on BraTS 2015. It has been also observed that Attention models tried to improvise on solving the problem of class imbalance in the brain tumor segmentation task in this model through sharing parameters via feedback from one model to another which reduced the problem attributed to multi stage CNNs to solve class imbalance on BraTS 2020 datasets. newlineWith recent advancements in Deep Neural Networks (DNN) for image classification tasks, intelligent medical image segmentation is an exciting direction for Brain Tumor research. DNN requires a lo