Comparative Performance Evaluation of VGG-19 and ResNet-50 on Brain MRI Images

DOI

https://www.doi.org/10.59256/ijsreat.20250506004

Published In

Volume 5, Issue 6 (2025)

Published Online

November-December 2025

Pages

25-31

Article Type

Original Article

Check for updates

The brain governs all bodily functions, making its integrity vital to human life. Brain tumors, which may be benign or malignant, represent abnormal growths within brain tissues and pose serious health threats. With a survival rate of approximately 75.2% for primary brain tumors, early diagnosis is essential. Traditional diagnostic methods are labor-intensive and prone to human error. To address these limitations, this study presents a deep learning-based approach for automated brain tumor detection using Magnetic Resonance Imaging (MRI). Specifically, two Convolutional Neural Network (CNN) architectures VGG19 and ResNet50 were evaluated on a brain MRI dataset categorized into four classes as “glioma”, “meningioma”, “pituitary”, and “notumor”. The VGG19 model attained an accuracy of (91%) with notable class-wise precision and F1-scores, while ResNet50 outperformed it with a remarkable (98%) accuracy and consistently high evaluation metrics across all classes. These results highlight the effectiveness of deep learning models, particularly ResNet50, in supporting computer-aided diagnosis for brain tumor classification.