DefinePK

Abstract

Doctors have a difficult time detecting a brain tumor at an early stage. Noise and other environmental disturbances are more common in MRI pictures. As a result, doctors have a tough time identifying malignancies and their origins. In the realm of medical image processing, a convolutional neural network is widely employed. Many academics have worked over the years to develop a model that can more accurately detect tumors. Here is an attempt to develop a model that can accurately classify tumors from 2D MRI scans of the brain. Although a fully connected neural network can detect tumors, we chose CNN, ResNet-50 and VGG-16 as our models due to parameter sharing and connection sparsity. These models will be incorporated to identify the type of tumor, such as a glioma, meningioma, or pituitary tumor, and recommend treatment options. Preprocessing is required to convert the image to grayscale. Filters are applied to photographs to reduce noise and other environmental influences. The image must be chosen/input by the user. The image will be processed by the system using image processing techniques. To detect malignancies in brain pictures, we used a deep learning algorithm. Convolutional Neural Network (CNN) which is implemented using Keras and Tensorflow and Pre-trained models such as ResNet-50 and VGG-16 are also built to compare the results