Akademik Veri Yönetim Sistemi
31st IEEE Conference on Signal Processing and Communications Applications, SIU 2023, İstanbul, Türkiye, 5 - 08 Temmuz 2023
Early diagnosis is very important in brain tumors. Although Magnetic Resonance (MRI) is widely used for brain tumor detection, it is difficult to detect the tumor manually. Therefore, computer-aided diagnosis systems have been frequently utilized in recent years. In this study, an Efficient Channel Attention-Dense Convolutional Network (ECA-DenseNet) framework is proposed to detect tumors in patients based on brain MRI images. While detecting the tumor, it is tried to determine which type of tumor is present in the patient. In the developed ECA-DenseNet structure, an ECA block has been added to the dense blocks. The ECA block aimed to discard unimportant information and thus reduce the computation time. The improved DenseNet model has been tested on an open-source dataset. The improved model is compared with DenseNet-121, DenseNet-169, DenseNet-201, and DenseNet-264. The experimental results show the improving model has better classification performance than the others. The accuracy of the proposed model was 95.07%.