@article{article_1535631, title={Alzheimer’s Disease Diagnosis in MRI Images Using Transfer Learning Methods: Evaluation of Different Model Performances}, journal={Balkan Journal of Electrical and Computer Engineering}, volume={13}, pages={119–127}, year={2025}, DOI={10.17694/bajece.1535631}, author={Gündüzvar, Eren and Kayık, Abdulsamet and Altuncu, Mehmet Ali}, keywords={Alzheimer, CNN architectures, MRI imaging, Transfer learning.}, abstract={One of the most prevalent types of dementia, Alzheimer’s disease, has become a serious health issue, especially among elderly individuals. Although there is currently no definitive cure for this disease, it is well known that patient care imposes substantial financial and psychological burdens on caregivers. As a result, early detection of Alzheimer’s disease is essential for stopping its progression and enhancing patients’ quality of life. Magnetic Resonance Imaging (MRI) is a widely used technique in clinical practice for diagnosing Alzheimer’s disease. In this study, various transfer learning models based on different CNN architectures, such as VGG-19, ResNet-50, DenseNet-201, and InceptionV3, were examined, and their performances were compared in detail to classify the stages of Alzheimer’s disease using MRI images. The models were tested on a publicly available dataset comprising four classes. The results demonstrate that the DenseNet-201 model, in particular, outperforms the other models in classifying the stages of Alzheimer’s disease.}, number={2}, publisher={MUSA YILMAZ}