Utilizing Deep Learning for Early Diagnosis of Dementia from Brain Scans and Patient characteristics

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder and the most prevalent form of dementia, primarily affecting middle-aged and elderly individuals. It is characterized by a gradual loss of cognitive abilities, memory loss, and behavioural changes severe enough to disrupt daily life. The disease is marked by the build-up of amyloid plaques and tau tangles in the brain, leading to neuronal death and cognitive decline. Although there is currently no cure for AD, research suggests a link between the risk of developing AD and various brain characteristics, such as grey matter atrophy and amyloid deposits. This research focused on comparing the performance of three machine learning models in identifying subtle, non-human-discernible features in brain scans to detect early indicators of Alzheimer’s. Using the OASIS-3 dataset, which includes MRI images and associated clinical data, this research compares the performance of three machine learning models in classifying cognitive states into four categories: Cognitively Normal, Questionable, Mild, and Moderate or Severe. The models evaluated were SoftMax Regression, Artificial Neural Network (ANN), and a hybrid Convolutional Neural Network (CNN). Each model's performance was assessed using key metrics such as precision, recall, F1 score, and confusion matrix. Among the models, the hybrid-CNN model achieved the highest accuracy of 98.41% and a precision of 98.77%, significantly outperforming the others. This comparative study highlights the potential of deep learning techniques in identifying complex patterns in brain scans that are often undetectable through conventional methods, offering valuable insights for early dementia diagnosis.