Improved Brain Tumor Detection Using Convolutional Neural Networks (CNN) and Transfer Learning Techniques for MRI Images

Brain tumours pose significant health challenges, necessitating accurate detection for effective treatment planning. This research explores deep learning architectures, including CNN, MobileNetV2, and EfficientNetBO, in brain tumour detection. Evaluated across glioma, meningioma, nontumor, and pituitary classes, EfficientNetBO outshines with a 99.16% test dataset accuracy, superior precision, and minimized errors. Its efficacy highlights transfer learning's dominance in medical imaging, leveraging pre-trained datasets for enhanced feature extraction. EfficientNetBO's reliability in brain tumor identification underscores transfer learning's significance, offering potential for improved medical diagnoses and treatments. Future research might delve into advanced techniques or practical integrations, promising enhanced brain tomour detection and treatment efficacy.