Automated Brain Tumor Detection and Comparative Analysis Using CNN Model and YOLOv8

The correct classification of brain tumors is an important factor in the early detection and treatment of neurological diseases. This paper suggests an automated deep learning-based system for brain tumor detection and classification based on CNN and YOLOv8. The data are MRI scans classified into four classes: Glioma, Meningioma, Pituitary Tumor, and No Tumor. Preprocessing methods, such as normalization and data augmentation, are utilized to improve generalization of the model. Hierarchical features from MRI images are extracted by the CNN model for accurate classification, and YOLOv8 is utilized for real-time detection of tumors. Accuracy, precision, recall, F1-score, and comparison with state-of-the-art approaches are used to evaluate performance. Interpretability methods such as Grad-Cam are also utilized to identify important regions affecting model predictions. Experimental results show that the CNN model has high classification accuracy, with YOLOv8 providing the best real-time detection. This work is an advancement toward enhancing automated tumor diagnosis and offers a robust tool for supporting clinical decision-making by radiologists. This work adds to the field of AI-based medical diagnosis by presenting a scalable and automatic computer-aided diagnosis (CAD) system for brain tumor classification. The designed system is of huge potential value in the field of reducing diagnostic time, enhancing accuracy, and enabling early detection, thus leading to improved patient care.