부산대학교 도서관

Brain tumors are dangerous malignant growths that require a precise diagnosis to be treated successfully. Traditional methods, such MRI scans and biopsies, usually rely on the expertise of medical specialists, which adds to the unpredictability and delays. Artificial neural networks (ANN) in particular has demonstrated promise for enhancing the accuracy of medical picture interpretation in deep learning. This work investigates the usage of ANN models for brain tumor identification and classification using sizable MRI datasets. In addition to assessing ANN topologies, the paper addresses problems with data pretreatment, model training, and validation techniques. The research findings aim to improve patient outcomes and treatment modalities by specifically establishing a deep learning framework with proven potential for early and precise brain tumor identification.