Alzheimer’s Disease Identification through Brain MRI Images: Leveraging Pretrained CNN Models

Alzheimer’s disease (AD) poses a significant global healthcare challenge, impacting an estimated 50 million individuals worldwide. As the leading cause of dementia, AD not only affects individual patients but also burdens families, caregivers, and healthcare systems. Moreover, the economic burden of AD is substantial, with healthcare and caregiving costs reaching hundreds of billions of dollars annually. Early detection of AD is crucial, as it allows for timely intervention and planning for patients and families. Recent advancements in medical imaging and machine learning offer promising avenues for early AD diagnosis. This study employs a methodology involving transfer learning and fine-tuning techniques with pretrained CNN models, including Xception, InceptionV3, VGG16, VGG19, and MobileNet, for AD detection using MRI brain images. The results demonstrate significant accuracy improvements during fine-tuning, with VGG19 achieving a perfect accuracy score of 100.00%. These findings underscore the urgency of developing accurate and efficient diagnostic tools for early AD detection, given the profound societal and economic implications of the disease.