Leveraging Handwriting Dynamics, Explainable AI and Machine Learning for Alzheimer Prediction

Abstract

Alzheimer’s Disease (AD) is a neurological condition with symptoms that gradually impair daily activities. Detecting it early, possibly up to 8 years before dementia symptoms appear, provides substantial benefits. Considering early detection as a motive, we leveraged the two-tiered stacking technique for efficiently classifying AD. We utilized the DARWIN dataset, specifically curated for AD identification through handwriting dynamics. Our proposed model combines predictions from various baseline classifiers at the initial stage, which is then used by a meta-classifier at the subsequent stage, amplifying the precision of diagnosis. Experimental trials revealed that our model boasts an impressive precision of 94.44% and other performance parameters. To augment transparency and understanding, we incorporated the principles of explainable AI, specifically using SHAP values—a state-of-the-art method to emphasize the features responsible for our model’s efficacy. Explainable AI (XAI) refers to methods and techniques in the field of artificial intelligence (AI) that make the results of AI systems more understandable to humans. Our proposed model holds promising potential for use in clinical applications. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.