Brain Tumors classification of MRI-Based Images using Deep Learning-based Approach

Brain tumors are created from the abnormal development of cells within the skull. Magnetic Resonance Imaging (MRI) is a significant healthcare modality. MRI images provide superior soft tissue contrast without ionizing radiation, unlike CT scans and X-rays including brain tumors. MRI’s positive impact is its ability to provide high-quality, non-invasive, and versatile diagnostic information, improving patient care and overall well-being. Before beginning treatment, the most significant problem is identifying and categorizing brain MRI scans. For a very long time, scientists have been trying to figure out the best method for more accurately identifying real-world medical images. Traditional approaches rely on manual interpretation, and feature extraction, making diagnosis time-consuming, error-prone, and subjective. Using a deep learning-based method can be automated with great accuracy and efficiency. This research investigates a customized Convolutional Neural Network (CNN) to classify brain tumors from MRI images. The experimental result of customized CNN obtained 99.57% total accuracy, 99.98% f1-score, 99.99% precision, and 99.98% recall respectively, which is more efficient than existing techniques for identifying brain tumors. According to this research, MRI is a state-of-the-art modality that can transform patient care by offering tailored diagnoses. Radiologists and doctors can obtain a more efficient and precise diagnosis process by utilizing the capabilities of this customized CNN.