This repository contains code for a deep learning model that detects brain tumors in MRI images. The model is implemented using a fine-tuned ResNet-50 architecture and trained on a dataset of 5,712 images, including Glioma, Meningioma, Pituitary, and Normal classes.
The dataset used for training the model consists of 1,321 images of Glioma, 1,339 images of Meningioma, 1,457 images of Pituitary tumors, and 1,595 images of normal/healthy brains. The images are in the MRI format and have been preprocessed for training the model. Pre-processed Dataset for both training and testing can be found here
The brain tumor detection model utilizes a fine-tuned ResNet-50 architecture. ResNet-50 is a deep convolutional neural network that has been pretrained on a large dataset and has shown excellent performance in various computer vision tasks.
The implementation of the model is done in Python using the TensorFlow framework. Ensure that you have the following dependencies installed:
Python 3.x TensorFlow Other necessary libraries (NumPy, Pandas, Matplotlib, etc.)
The model has been trained to achieve high accuracy in classifying brain MRI images into different tumor types and normal/healthy categories. The accuracy and performance of the model can be evaluated by testing it on an independent test dataset or real-world MRI images.