A Review on Magnetic Resonance Spectroscopy for Clinical Diagnosis of Brain Tumour using Deep Learning

Medical imaging is a way to analyze the anatomical structure of the body. With the advent of new technologies, more and more sophisticated imaging techniques have been developed such as MRI, CT, PET, X-ray, etc. which have made the non-invasive diagnosis of diseases a lot easier. Magnetic Resonance Spectroscopy (MRS) gives detailed information about the chemical changes occurring in the brain by measuring the change of metabolites. It can be used to diagnose tumors in the brain and other brain-related diseases and disorders. Magnetic resonance spectroscopy is an encouraging tool that can be used for the effective diagnosis and grading of brain tumors but it suffers from many problems such as baseline distortions and artifacts, low resolution, low intensity images which makes its applicability a practical difficulty. The advancements in the field of Deep Learning can be effectively used to tackle the challenges in MRS which in turn can be employed for the brain tumor detection and classification.