Abstract

Accurate and timely classification of brain tumors is critical for effective diagnosis and treatment planning. In this study, we propose a multiscale deep learning model that leverages hierarchical spatial features from magnetic resonance imaging (MRI) scans to automatically classify brain tumors. The model integrates multiple convolutional neural network (CNN) branches operating at different image resolutions to capture both global contextual information and fine-grained local details. We trained and validated our model on a publicly available brain tumor dataset, including glioma, meningioma, and pituitary tumor classes. Performance evaluation using metrics such as accuracy, precision, recall, F1-score, and AUC-ROC demonstrated that the multiscale approach significantly outperforms single-scale models. Our results suggest that multiscale deep learning offers a robust and scalable solution for clinical decision support systems in neuro-oncology.