Current Trends on Deep Learning Models for Brain Tumor Segmentation and Detection – A Review

Abstract

Critical component in diagnosing tumor, designing treatment and developing an outcome for evaluating brain tumor segmentation needed to be highly accurate and reliable. Magnetic Resonance Imaging (MRI) help and support the health care field to detect the very minor abnormal growth in any part of the human being. While deep neural networks (NNs) and machine learning techniques have good achievement in 2D image segmentations, but it's a challenging task for NNs to segment critical organs from 3D medical MR images. Segmentation relating tumor detection includes several processing techniques that are categorized into Pre-Processing, Segmentation, Optimization and Feature Extraction. Study focuses mainly on 3D-based Convolution neural network (CNN), ANN (Artificial Neural Networks), SVM and Multi-class Support vector machines (MCSVM) for Deeper Segmentation. To remove computational burden of processing 3D medical scans, this survey paper plan to review the current development in image segmentation and image classification based on efficient and effective towards processing of tumor infected human brain MRI adjacent image patches that can pass through the network with a target on gliomas, while robotically adapting towards an imbalance present in the data. Thus, more discriminative 3D NNs and Computational Machine learning that assists in processing the input images at multiple scales simultaneously. Finally, this article implying about present status on segmentation and Detection of tumor-based image processing through deep learning models.