Abstract

Brain tumor detection and classification remain critical challenges in the medical field, necessitating precise and efficient diagnostic tools. This project presents an advanced approach for brain tumor prediction using deep learning techniques, specifically leveraging the ResNet50 algorithm to classify brain tumors into four distinct categories: glioma, meningioma, pituitary tumor, and no tumor.

The base IEEE paper outlines the effectiveness of CNNs in detecting the presence of tumors. However, our approach enhances this by employing the ResNet50 architecture which allows more robust feature extraction. In addition to detecting only presence, this project detects different categories of brain tumor such as glioma, meningioma, pituitary tumor, and no tumor.

The dataset used in this project comprises MRI scans. The ResNet50 model was fine-tuned on this dataset, achieving significant accuracy improvements in classifying the four tumor types.

You can download abstract from here

Technologies Used

Modules

1. User

• 🧠 Login/Signup
• 🧠 Upload MRI images
• 🧠 4 Classes - Glioma tumor, Meningioma tumor, pituitary tumor and No tumor
• 🧠 View History of previous predictions