Brain Tumor Segmentation and Survival Prediction using Multimodal MRI Scans with Deep learning Algorithms

Abstract

Brain tumours developed by the brain’s abnormal cell growth. Medical imaging devices are classified into two types: MRI and CT are widely used to scan brain tumors. MRI images are used to scan the interior components of the brain. There are two types of brain tumours: benign and malignant. A benign brain tumour is one that can be cured, whereas a malignant brain tumour is one that cannot be cured. Gliomas and astrocytomas are two types of malignant brain tumours. A radiologist using conventional procedures to detect and diagnose a brain tumour. Errors and delays are all too common. Because neurosurgeons generate a large number of brain tumour images, imaging technicians are unable to manually categorise and segment them. It is not suitable for densely populated emerging markets. As a result, computer-aided automatic brain tumour detection and diagnosis is preferred. Learning algorithms, such as neural networks, are hampered by a lack of hyperparameter modification, several local optima, and higher computing time. Optimizing network architecture and hyperparameters is critical for avoiding these challenges. Deep learning optimization approaches are commonly employed in healthcare. This research proposed using Deep Learning Optimization Algorithms to detect and diagnose Meningioma brain cancer. This intended study addresses the extraction, categorization, and diagnosis of tumor picture segments.