Unlocking Alzheimer's: A Squeezenet-Based Approach for Automated Diagnosis Across Disease Stages

The neurodegenerative disorder known as Alzheimer's Disease (AD), which is carried on by the death of brain cells are characterised by memory loss and cognitive decline. The complexity of the brain's structure and functioning makes early identification of AD challenging, despite the fact that research into the illness has considerably increased recently. In this work, Multiclass Classification of Alzheimers disease stages using Squeeznet based approach for automated diagnosis is presented. This work mainly focused on multiple stages of AD classification. For identifying the phases, including Mild Cognitive Impairment (MCI), EMCI (Early MCI), Late MCI (LMCI), and CN (Cognitively Normal), the SqueezeNet with Harris Hawks Optimisation approach (HHO-SqueezeNet) is described. These feature vectors are processed using the MRI, and the aforementioned iTask is subjected to the PCA technique for the feature dimension. Additionally, we employ the CML-ELN strategy for each job in order to account for the quantity of related tasks. As a result, only performance metrics such as model accuracy, sensitivity, specificity, precision, and recall that may be pertinent to the issue are employed. It will be feasible to classify the appropriate subjects. With relation to SqueezeNet, the MTL algorithm created in this work is quicker, more effective, and most like an AD treatment.