Brain Tumor Classification Using Deep Learning Techniques

This research paper provides an in-depth analysis of the recent advancements in the use of deep learning techniques for the detection and classification of brain tumors into three main categories: benign, malignant, and pituitary. A comprehensive analysis of the major research papers in this field is presented, covering the different deep learning architectures and techniques used, and their performance in detecting and classifying brain tumors with an accuracy range of 90–99%. The paper highlights the potential of deep learning in improving the accuracy and speed of brain tumor detection and classification in an attempt to reduce the use of biopsy, while also identifying existing limitations and prospects for future work. Additionally, this review provides a comparative study of different methods that can be used and their performance in detecting and classifying brain tumors into the above-mentioned tumor types. Accuracy results obtained from CNN traditional approaches were in the range of 85.62–96.65% and that of experimental approaches from 91.44–96.36%.