Can we rely on AI or Deep Learning to read or detect brain tumors using MRI scan and images?

Abstract

The brain is one of the complex and most significant organs in the human body. It controls every process that regulates our body and affects how we think, feel, act and is indispensable to living a happy and balanced life. Due to the current lifestyle, depression, stress, and anxiety have risen to an extreme, especially during COVID. Researchers have found that people with mental disorders are more prone to brain tumors and develop cancers early than people without mental illness. Therefore, it is worthwhile to invest time and research to improve our ability to diagnose brain tumor in early stage. This research aims to use the Deep Learning model to detect brain tumors. The dataset used for this research project is the Brain MRI segmentation dataset (LGG lower-grade gliomas) taken from The Cancer Imaging Archive (TCIA). This dataset is open source and available on Kaggle. Using the Keras library, I used the Convolutional Neural Network (CNN) for tumor detection to get concrete results. I got approximately a 90% accuracy which shows us it is very promising that Deep Learning can help with brain tumor predictions, thereby making the diagnosis easy, correct, and prompt, which allows early diagnosis and reduces the mortality rate.