Comprehensive Analysis of Deep Learning Models for Brain Tumor Detection from Medical Imaging

Deep learning is a powerful tool of machine learning has proven remarkable performance in many domains like medical science, defense, climate analysis, and many more. Deep learning models with computer vision methods have many applications in medical sector to detect and analyze various diseases from images and videos. Medical sector images like X-ray, CT scan, MRI, and others are mostly used to detect spreading of disease from human organs. Proposed study uses comprehensive analysis of brain tumor detection over various deep learning models like VGG16, AlexNet, ResNet50, GoogleNet, and DenseNet121. Proposed study analyzes and describes different convolution methods for tumor classification. Proposed study also explores the impact of various training parameters like train-test split ratio, dropout, and others. Simulation of proposed analysis used Br35H and BraTs 2017 dataset for binary classification. Proposed study has also analyzed and compared various states of the art classifier to find superior performing model of computer vision. Proposed analyzed study finds DensNet121 with MLP classifier has achieved 97.86% remarkable accuracy.