MRI-Based Deep Learning Classification of Alzheimer’s and Parkinson’s Disease
摘要
This study investigates the potential of transfer learning, evaluates the impact of different deep learning designs on classification performance, and tests whether deep learning models can accurately classify Alzheimer’s Disease (AD) and Parkinson’s Disease (PD) from MRI images. This study employs experimental analysis for data collection, preprocessing, and deep learning model training. We used VGG, ResNet and EfficientNet-B4. The result using EfficientNet-B4 giving an accuracy of 98% demonstrates the effectiveness of deep learning in precisely identifying AD and PD, providing insights for better diagnosis and patient care.