Utilizing Novel Convolutional Neural Networks for The Detection of Brain Tumors in MRI Images

Abstract

Embarking on a critical exploration of brain tumor detection, our research delves into the intricacies of this field, acknowledging the vital need for precise diagnostics. Employing custom Convolutional Neural Network (CNN) models, we aim to surpass traditional approaches by capturing the nuanced details of brain tumors. With a significant toll from malignant brain tumors each year, the imperative for accurate diagnostic tools is unmistakable. Motivated not just by technical considerations, but by a deep commitment to the profound human impact of brain tumors, our research seeks to contribute to more effective diagnostic tools. We aspire to provide adaptability to the unique characteristics of these growths, offering potential societal benefits such as timely interventions, personalized treatment plans, and improved prognoses. This study represents a dedicated response to the collective call to address the challenges posed by brain tumors, leveraging technology for a meaningful impact on individual lives and broader societal well-being. The proposed custom CNN model outperformed traditional transfer learning models, achieving an impressive overall accuracy of 96.30%.