Akademik Veri Yönetim Sistemi
16TH INTERNATIONAL İSTANBUL SCIENTIFIC RESEARCH CONGRESS ON LIFE, ENGINEERING, ARCHITECTURE, AND MATHEMATICAL SCIENCES PROCEEDINGS BOOK, İstanbul, Türkiye, 28 - 29 Şubat 2024, ss.376-383
Brain and other central nervous system (CNS) tumors are among the deadliest cancers. It is a common type of cancer in adults and children. Detection of the disease at an early stage and correct classification of the tumor significantly increases the survival rate. For this reason, medical image analysis is an essential field of study for cancer detection and classification. In this paper, image processing and deep learning methods are used to analyze brain tumor magnetic resonance imaging (MRI) images from the BR35H dataset. The data set we used in this study contains 2 classes, these classes contain tumor and non-tumor MRI images. The MR images in this dataset were resized and data augmentation algorithms were applied. Feature Extraction is the act of reducing the size of image data by extracting the relevant information from the image. In this study, we extracted texture features using Grey Level Co-occurrence Matrix (GLCM) and shape features using Canny Edge Detection and applied them to Inception-V3 image classification model, which gives 99.54% accuracy rate. When the experimental results of the suggested approach were examined, satisfactory performance was observed. Thus, it was concluded that the proposed method can be used for computer-aided decision making systems in the clinical diagnosis of brain tumors.