Deep Learning in Medical Image Classification from MRI-based Brain Tumor Images

TL;DR

This study tackles four-class brain tumor classification from MRI scans and compares four pretrained CNNs (MobileNetV2, EfficientNet-B0, ResNet-18, VGG16) with a novel MobileNet-BT architecture. By leveraging transfer learning, data augmentation, and a fine-tuned classifier, MobileNet-BT achieves superior accuracy and F1-score (0.9924) on a Brain Tumor MRI Dataset of 7023 images, surpassing the best pretrained model (VGG16) at 0.9497. The work demonstrates that unfreezing the entire backbone and adopting a custom classifier can substantially improve performance for medical imaging tasks while reducing training epochs. These findings suggest that careful architectural customization is essential when transferring large CNNs to domain-specific medical image classification problems, enabling more reliable and efficient decision support.

Abstract

Brain tumors are among the deadliest diseases in the world. Magnetic Resonance Imaging (MRI) is one of the most effective ways to detect brain tumors. Accurate detection of brain tumors based on MRI scans is critical, as it can potentially save many lives and facilitate better decision-making at the early stages of the disease. Within our paper, four different types of MRI-based images have been collected from the database: glioma tumor, no tumor, pituitary tumor, and meningioma tumor. Our study focuses on making predictions for brain tumor classification. Five models, including four pre-trained models (MobileNet, EfficientNet-B0, ResNet-18, and VGG16) and one new model, MobileNet-BT, have been proposed for this study.

Figures (6)

• Figure 1: Brain Tumor Types
• Figure 2: EfficientNetB0 Loss VS Accuracy
• Figure 3: MobileNetV2 Loss VS Accuracy
• Figure 4: Resnet18 Loss VS Accuracy
• Figure 5: VGG16 Loss VS Accuracy