Performance comparison of multiple neural networks for brain tumor classification

Malignant brain tumors impose physical, emotional, and societal burdens on patients. Common symptoms, such as headaches, seizures, vision problems, personality changes, and cognitive impairments, highlight the urgent need for early diagnosis, effective therapies, and precise prognosis. Over the past decade, medical applications of deep learning have made significant progress, particularly in tumor detection and diagnosis. However, given the intricate nature of medical data, a comprehensive evaluation of the strengths and weaknesses of deep learning methods and models is imperative to optimize their utility. Four canonical deep learning models, VGG16, VGG19, ResNet50, and Xception, were trained and tested for the classification of MRI images of three prevalent brain tumors: glioma, meningioma, and pituitary tumor. The results reveal that Xception achieved the highest classification accuracy (0.97), demonstrating its immense potential for diverse and complex medical applications. The results presented herein serve as a valuable reference for selecting deep learning models tailored to specific conditions and tasks, thus facilitating more informed decision-making in the field of medical imaging and contributing to the broader goal of enhancing brain tumor management.