İleri Mühendislik Çalışmaları ve Teknolojileri Dergisi
Yazarlar: Muhammed Oguz TAS, Semih ERGİN
Konular:Mühendislik
Anahtar Kelimeler:Magnetic resonance imaging,Brain tumor detection,Brain tumor segmentation,Deep learning,K-means
Özet: Tumors occur in the brain as the cells in the brain tissue grow abnormally. Since a large amount of tumors in the brain are cancerous, it can have consequences until the death of the sick person. MR imaging is widely used as a means of imaging brain tumors. MR images can distinguish diseased areas and healthy areas using the image's texture, contrast, brightness and boundary information. In this way, planning of the treatment process of the disease can be made by finding the shape, location, size and area of the brain tumor. In this study, the detection of the brain tumor in MR images by using deep learning and the segmentation with K-means are performed. As a result of the study, the accuracy obtained in detecting of the brain tumor is 84.45%, and the sensitivity is 95.04%. The study proposed detection and segmentation of the brain tumor and, extracting the tumor area automatically.