Brain Tumour Segmentation Using Wasserstein Generative Adversarial Networks(WGANs)

The majority of health-related applications use technology and require a lot of data to operate. Even more, information is required for brain tumor segmentation, but many people lack access to it because of privacy laws governing medical information. Therefore, this study utilizes GAN technology, which creates synthetic images to resolve this issue. AGGrGAN is used for aggregation and discrimination, and PGAN, WGAN, and DCGAN are used to create synthetic pictures. Synthetic brain tumor masks with the same visual characteristics as the real samples. potential benefits of using GANs to synthesize new images from a real dataset of brain MR scans and corresponding hand-labelled segmentation masks (annotations), and an evaluation of the potential performance gains that are achieved when using synthetic images to augment a dataset that is used to train a neural network to solve a brain tumour segmentation task.

I. Introduction

Artificial intelligence (AI) has created several opportunities to raise our standard of living. The advantages of AI also extend to healthcare facilities and medical diagnoses. It has been demonstrated that a variety of deep learning approaches are useful for Health visual analysis, this also improves understanding of illnesses that impact the body of the human. Medical imaging includes the field of brain tumour scans. Tumors develop as a result of uncontrollable cell development in specific areas of the body of the human, brain. Early-stage brain tumours must be properly identified in order to prevent death, while more severe cases of brain tumours may result in long-term damage in our bodies. In the previous few years, the India has recorded an enormous number of cases of brain tumours, many of which resulted in death [1]. As it involves life and death, numerous studies have used various deep learning and machine learning technologies to segment brain tumors. To get brain scan we often use MRI which is Magnetic Resonance Imaging used to scan cell tissues and organs to identify the issue, MRI uses magnetic waves for finding any Tumor's in the brain. The current tumor's size can also be predicted by MRI. Due to its overall efficacy, MRI is chosen over other procedures for detecting brain tumours, even though it has drawbacks such taking too much time or triggering Claustrophobia. As a potent non-invasive analytical technique, MRI is very helpful to make diagnoses using images. We have a few publically accessible datasets for research purposes, but they are sparse and do not contain enough information about other health problems. Researchers are unable to access certain datasets due to patient data protection [2].