MemorySona: Illuminating Cognitive Health with Deep Learning - A Mobile Medical App for Alzheimer's Patients, Emphasizing Detection through Brain MRI Images

MemorySona is a recent development and specialized application for Alzheimer's patients. It is designed for individuals with Alzheimer's Disease (AD) to provide support in various ways, such as cognitive training, memory exercises, reminders, and communication assistance. It can be part of a broader approach to improve the quality of life for individuals with Alzheimer's and support their caregivers. It follows ethical guidelines for patient privacy and data security. “MemorySona” is a unique and creative name for a memory app. It combines “Memory” and “Sona” to create a distinctive and memorable title. The “Sona” is a concise and elegant name that could work well for a memory or mindfulness app. It suggests a harmonious or beneficial aspect of memory enhancement as it has a simple and soothing sound, which aligns with the idea of enhancing memory or promoting mental well-being. Depending on the specific focus and features of the app., “Sona” could be an excellent choice. The primary objectives of MemorySona include achieving accurate classification outcomes through the utilization of freely accessible data and maximizing accuracy across a vast collection of Brain MRI Images. This is accomplished by assessing the performance of five deep learning models on identical datasets to determine the most effective approach. The research uses five deep learning models, CNN, DensNet121, ResNet50, VGG16, and Inception V3 to classify brain MIR images for Alzheimer's detection. According to the obtained results, DnseNet121 has the height evaluation by 89.50%, 89.96%, 97.79%, 90.03%, and 88.36% for recall, precision, Auc, F1-score, and accuracy respectively.