Multi-Class Brain Tumor Classification Using GAN-Enhanced Deep Convolutional Networks

Abstract

Brain tumors are considered one of the deadliest neurological conditions, where timely and precise diagnosis is crucial for improving survival rates and effective medical treatment. This work presents a hybrid deep learning–driven framework for multi-class brain tumor classification. The proposed approach integrates a deep convolutional generative adversarial network (DCGAN) with state-of-the-art convolutional neural network-based classification models, namely EfficientNetB3, DenseNet201, InceptionResNetV2 and Xception. In addition, the quality of the synthesized MRI images is evaluated using a range of performance metrics, namely Fréchet Inception Distance, peak signal-to-noise ratio, structural similarity index measure and Inception Score. Among all evaluated models, the DCGAN-enhanced EfficientNetB3 achieved the highest classification performance, with an accuracy of 95.20%, precision of 94.87%, recall of 94.52% and an F1-score of 94.69%.