- Beatriz Vizoso
- Filipa Alves
- Helena Oliveira
- Maria Almeida
Master in Data Science and Advanced Analytics (Nova IMS, Lisbon) - Spring 2022
This project focuses on the classification of types of brain tumors given an MRI image of the brain, by designing and implementing a Deep Learning model. The dataset used contains 3064 contrast enhanced images from 233 patients with 3 kinds of brain tumors: meningioma, glioma and pituitary.
The goal was to develop a Deep Learning model that could input MRI images of patients with one of the previously mentioned tumors and be able to identify which of those tumors each patient was suffering from based solely on that image, making it a multiclass classification problem. This problem can have very important real-life implementations, as correct detection of brain tumors will influence the treatments given to the patient and, consequently, influence their recovery.