Enhanced Detection of Brain Tumors in MRI Scans via Convolutional Neural Networks

The treatment of brain tumours, particularly in advanced stages, presents significant challenges due to its high risk and substantial costs. Historically, the identification of such tumours relied heavily on manual diagnosis, which, while effective, was prone to human error and required considerable labour. The advent of artificial intelligence has introduced advanced detection tools that can significantly enhance diagnostic accuracy of which, CNNs have demonstrated exceptional proficiency in recognizing patterns within visual data. This paper proposes a model that leverages CNNs to classify MRI scans into tumorous or non-tumorous categories. Our approach aims to improve early detection and diagnosis by utilizing the CNN's capability to analyse complex patterns in imaging data with high precision and to learn and adapt from new data. We detail the architecture of our CNN model, the dataset that is being used for training and also for validation, and the metrics of evaluation are employed to evaluate the performance. Our research's findings show that the suggested model attains a high degree of accuracy in distinguishing between tumorous and non-tumorous scans, making it a viable tool in diagnostic settings.