Enhanced magnetic resonance imaging feature extraction for precise brain tumor classification using dual deep convolutional networks

Denis Bernard, Constantino Msigwa, Jaeseok Yun

Abstract: Highlights•We propose the Dual Deep Convolutional Brain Tumor Network (D2CBTN) to enhance feature extraction and improve brain tumor classification accuracy.•We implemented four convolutional neural network-based transfer learning algorithms to investigate their capabilities for brain tumor classification.•We explored the Vision Transformer and Swin Transformer models for brain tumor classification.•We employed a data augmentation technique (i.e., ImageDataGenerator function) to address data imbalance and improve the accuracy of brain tumor classification.•The experimental results of D2CBTN demonstrate superior performance compared to existing state-of-the-art brain tumor classification techniques.

External IDs:doi:10.1016/j.knosys.2025.114628