| dc.description.abstract | Alzheimer's Disease (AD) is a gradually degenerative disease that affects the brain and nervous system. Due to the global increase in the aging process, the incidence of Alzheimer's disease has been increasing every year. Although this disease is very common, there is currently no effective cure for Alzheimer's disease. Therefore, early detection and accurate diagnosis in the early stages are very important for preventing more serious stages and for further patient care and treatment. Doctors can detect Alzheimer's disease by looking at images obtained from Magnetic Resonance Imaging (MRI) of the brain with the naked eye. However, human error may occur. In addition, manual evaluation of brain MRI can be very time-consuming and prone to errors due to variable errors. To reduce these errors and improve effectiveness, a system is needed that can automatically predict and detect various stages of Alzheimer's disease. In this study, the EfficientDet method is used to classify Alzheimer's disease into five classes, namely: Non Demented, Very Mild Demented, Mild Demented, Moderate Demented, and Severe Demented. Before the classification process, image data undergoes pre-processing, including resizing, grayscaling, and augmentation. Testing was carried out on 196 brain MRI images using a web-based system. Based on the results of the testing, the system was able to classify Alzheimer's disease with a precision value of 0.977, recall of 0.979, F1-score of 0.977, and accuracy of 97.4%. | en_US |
