Brain Tumour Detection Using Convolutional Neural Networks in Machine Learning: A Streamlit-Based Framework for MRI Image Analysis

Abstract

Objectives: The deep learning model's capacity to recognise brain cancers in MRI images. The model is intended to automatically analyse MRI scans and determine whether a tumour is present, producing reliable, accurate classification results. Methods: A previously trained CNN was used to classify MRI images. The MRI image in the dataset was first reduced and normalised during preprocessing to ensure accurate input to the model. After processing, the model produced a probability value indicating the likelihood that the image contained a tumour. Findings: The available MRI cases were correctly classified as tumours, and no non-tumour cases were identified. The prediction's tumour probability of 99.74% indicates how confident the model was in its classification result. Novelty: This work demonstrates a CNN-based approach to identifying brain tumours from MRI data. Even with a small input sample, the system generated accurate and reliable predictions. The proposed method demonstrates how deep learning models could aid in cancer detection and potentially serve as a useful adjunct to clinical decision-making.