Alzheimer’s disease: Early-Stage Prediction and Classification using Multi-Model Technique

Alzheimer’s Disease(AD) is a neurologic confusion with no cure. The first phases of disease prediction help to prevent and control further growth. The focus of this research is the initial stage of classification and prediction of AD. This work produces quick solutions for the initial stage of prediction and before going to the next stage of prediction. To predict the early stages of AD, various methodologies and criteria, as well as various machine learning algorithms are used. But still, the main research gap is that the level of prediction and accuracy does not reach the mark. So, in this work, multi-model techniques for the early-stage prediction and classification are introduced. This work suggests a novel hybrid method based on biomarkers, static features, dynamic features, and convolution neural networks with hyperparameter tuning (CN-HP) that helps with initial prediction of early stages of AD. The proposed CN-HP method is evaluated using different metrics such as accuracy, specificity, and sensitivity. The work is then compared with traditional prediction models and corresponding metrics for better results, and also the comparison result has produced better accuracy.