Brain Tumor Classification Using Convolutional Neural Networks

Abstract

Delays in detecting brain tumors can result in worsening health and a greater financial strain on healthcare systems. Brain tumors are a very significant and complicated issue. This work suggests a methodology for the early detection of brain tumors with MRI pictures, which could lead to greater diagnosis accuracy and lower treatment costs in the later stages. The MRI pictures where the classification is divided into four groups: gliomas, meningiomas, pituitary tumors, and non-tumor cases. Advanced techniques including picture enhancement and data augmentation were used to boost model performance and correct dataset imbalances. The findings showed an astounding accuracy of 99.69%, proving the model superiority to current models.