Explainable Artificial Intelligence Reveals Shared Pathogenetic Mechanisms Between Alzheimer’s and Parkinson’s Diseases: Development of Gender-Specific and Non-Invasive Diagnostic Systems

Introduction. The Neurodegenerative Elderly Syndrome (NES) hypothesis proposes that Alzheimer’s disease (AD) and Parkinson’s disease (PD) share early pathogenic mechanisms, including monoaminergic dysfunction and alpha-synuclein accumulation. Studying AD and PD within a unified framework may enable the identification of biomarkers at stages in which neurodegenerative trajectories are not yet clinically differentiated. Aim: To validate the NES paradigm through Explainable Artificial Intelligence (XAI), identifying gender-specific predictors and developing a non-invasive clinical decision support system. Methods: Machine learning models were trained on longitudinal datasets integrating clinical and neuropsychological variables. Explainable AI techniques were used to quantify predictor importance with a gender-sensitive approach. The resulting models were implemented in the Web App eXplAIn Medical Analysis (EMA), co-designed with clinicians to provide interpretable diagnostic profiles based exclusively on non-invasive measures. Results: Findings support partially shared diagnostic pathways between AD and PD, alongside gender-related differences in predictor relevance (e.g., verbal memory performance, education level). The EMA platform, currently focused on AD, demonstrated high accuracy in predicting clinical conversion up to 18 months before standard diagnosis. Discussion: The NES framework helps overcome current diagnostic fragmentation by promoting a pathophysiology-oriented perspective on neurodegeneration. By integrating Explainable AI, the approach ensures clinical transparency and interpretability, supporting an augmented medicine paradigm in which AI enhances—rather than replaces—clinical reasoning. Conclusion: The proposed strategy provides a non-invasive, gender-sensitive solution for early Alzheimer’s disease detection and establishes the scientific foundation for extending the system to Parkinson’s disease. This framework fosters a human–machine synergy aimed at advancing personalized and precision-based neurodegenerative care.