Brain Tumor Detection From MRI Images Using CNN-Based Deep Learning Models

Authors: Rishi Kumar Mishra, MD Yasin Alam, Shivam Pisudde, Mohammed Abubakar

Abstract: Early and accurate detection of brain tumors from magnetic resonance imaging (MRI) is critical for patient care. This paper presents a CNN-based pipeline for binary brain tumor detection using grayscale MRI images, built on a transfer-learning backbone (EfficientNetB0) with targeted preprocessing, augmentation, and explainability via Grad-CAM. We describe dataset handling, model architecture, training strategy, and evaluation metrics including accuracy, AUC, precision, recall and confusion analysis. Empirical results on commonly used MRI image collections demonstrate that the proposed workflow achieves competitive performance while remaining computationally efficient. We conclude with a discussion of limitations, reproducibility practices, and recommended future extensions.