KLASIFIKASI CITRA PENYAKIT TUMOR OTAK MRI MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK MODEL EFFICIENTNETV2B0

Brain tumors are abnormal cell growths within the brain that can affect the function of the nervous system. Early detection and classification of tumor types are very important to determine the right treatment. This study aims to classify the types of brain tumors using MRI images by applying the EfficientNetV2B0 Convolutional Neural Network (CNN) model. The dataset used in this study is "Brain Tumor Classification (MRI)" from the Kaggle website, which consists of four classes: no tumor, glioma, meningioma, and pituitary tumor. In this study, to increase the variety of training data, data augmentation techniques are used, which include rotation, flipping, and zooming of MRI images, and the EfficientNetV2B0 pretrained model is used because it has an efficient architecture and is good at the task of classifying images. The results show that the proposed model is able to classify brain tumor types with a high degree of accuracy. The EfficientNetV2B0 model achieved 97% accuracy with evaluation results using data test.