A Low-weight Deep Learning-Based Approach for Brain Tumor Classification

Brain tumors (BTs) are rapidly increasing in prevalence globally, resulting in thousands of fatalities each year. Accurate detection and classification are crucial for effective treatment. Various research methodologies have been developed for BT detection and classification. deep learning (DL) approaches excel in feature extraction and have gained significant traction for classification and detection tasks. In this study, we introduce a low-weight deep learning model, designed for the classification of three types of brain tumors: glioma, meningioma, and pituitary tumors. This model is based on a modified convolutional neural network (CNN) architecture. To enhance the model, we used MobileNet as our baseline model and incorporated the ELU activation function to improve the model's expressiveness. The performance of the proposed model was evaluated using a publicly available dataset, achieving impressive results with 97.67% accuracy, 97.6% precision, 97% recall, and a 97.66% F1-score. The findings demonstrate that our model is superior to existing models for classifying brain tumors from MRI images.

A Low-weight Deep Learning-Based Approach for Brain Tumor Classification Keywords:

deep learning , brain tumor , MRI , transfer learning , convolutional neural network