<![CDATA[Brain tumor is a highly dangerous and deadly disease. It can occur due to the abnormal growth of cells or tissues in the head. Treatment for brain tumor is done with surgery and chemotherapy aimed at killing or destroying the cells that affect the growth process of brain tumor. Diagnosis of brain tumor is done using medical scans such as MRI, CT Scan, and PET Scan by analyzing the resulting images. Another method used to detect brain tumors is through biopsy, which is a process of taking cells or tissue from the body for examination in the laboratory. However, this method takes a long time because the cells taken from the patient will be examined in the laboratory. Therefore, a technique is needed to speed up accurate brain tumor diagnosis in order to obtain quick treatment. Machine learning can solve this problem with the classification of images produced by MRI. The classification technique that can be used is the GoogLeNet architecture in CNN. Because GoogLeNet is the algorithm that won the ImageNet Large Scale Visual Recognition Challenge (ILSVC) in 2014 The purpose of this study is to classify brain images using the GoogLeNet architecture. The dataset used in this study consists of 7023 images, consisting of 6320 images for training the model and 703 for testing the model. The results of this study obtained an accuracy percentage of 96%. This result is higher than previous studies that obtained an accuracy value of 94%.]]>
Copyrights © 2024
