A Comprehensive Survey on MRI Images Classification for Brain Tumor Identification using Deep Learning Techniques

A brain tumor, a disease that can be fatal, threatens the most valuable human life, and it is a difficult work for a doctor to diagnose the tumor accurately and promptly. The aberrant proliferation of brain cells results in a condition known as a brain tumor. The rarity and variety of tumors make it difficult to gauge a patient’s prognosis after being diagnosed with one. Manual identification is a time-consuming and difficult method that can lead to inaccuracies in the results of tumor identification using Magnetic Resonance Imaging (MRI). MRI images play a vital role in tumor site determination. These limits necessitate the use of computer-assisted techniques. It is common practice to utilize MRI scans to identify a variety of tissue abnormalities, to look for tumors, and to assess whether a tumor is still present or returning. Deep learning (DL) algorithms are being utilized in neuroimaging to detect brain cancers using MR images as artificial intelligence advances. It is critical that medical photographs be processed to aid in the identification of various disorders. The information and expertise of the physician are critical in the diagnosis of brain tumors. Physicians need an automated method to detect and classify brain cancers. MRI-based segmentation of brain approaches will be reviewed in this research. Deep learning methods for automatic segmentation have recently gained popularity since they produce cutting-edge results and are more suited to dealing with this challenge. It is also possible to use deep learning approaches to analyses and evaluate vast quantities of MRI-based image data quickly and objectively. There is a slew of review studies about classic MRI-based approaches for classifying brain tumor pictures. Deep learning approaches were used to classify brain cancers as glioma, meningioma, or pituitary. Conclusions and future advances are also discussed in this section to ensure that MRI-based tumor segmentation methods can be implemented in daily practice.