Deep Learning-Based Diagnostic Tool for Brain Tumours

In this work, we provide the results of our experiments on Brain Tumour Classification and explore their implications. MRI, short for magnetic resonance imaging, is a crucial piece of medical imaging technology that creates high-resolution pictures of the inside of a human body by combining radio waves, a strong magnetic field, and a computer. In this study, we employ a CNN-based Deep learning model to categorise cases of brain tumours. Common applications for CNNs (Convolutional Neural Networks) include video and image analysis. It processes images by first ingesting them and then convoluting them into smaller and smaller pieces. The convolutional layer analyses the input picture using a series of filters designed to identify elements like edges, curves, and patterns. Here, we constructed the CNN using data obtained from Kaggle. It contains roughly 3300 MRI scans of the brain, split into 4 groups (no tumours, meningioma tumours, glioma tumours, and pituitary tumours) and a total of 384 testing pictures. We used Keras to construct the model. The accuracy of the suggested project is quite high, coming in at 97.142%. The Confusion matrix is used to evaluate the precision. It is a data table for measuring the efficacy of an ML system. This provides the results of a comparison between the actual values in a dataset and the model's projected values in an easy-to-understand format.